UPC_CFI_2594/2026; UPC_CFI_2595/2026 – Versuni v Dreame

Expert Commentary

Provisional measures Facts 1. This order concerns two preliminary injunction (“PI”) proceedings regarding EP 3 169 210 and EP 4 274 458. 2. The two cases were held together. They relate to a cleaning device for surfaces (a mop). The Court 1. The Court cites the principle (“more likely than not”) for obtaining a PI. 2. The Court decides that the claimant acted diligently and that urgency is given. 3. With respect to EP 210, the Court cites the problem the patent states to resolve and the core of the solution. 4. The Court cites the principles of claim interpretation and defines the skilled person. 5. The Court does not accept Dreame’s argument that “surface cleaning device” can also include “medical absorbing arrangements”. 6. The Court does, on the other hand, also not accept Dreame’s interpretation that the patent is limited to liquid removal. The same is true for all other limiting interpretations of the claim defended by Dreame. 7. The Court considers, with respect to novelty, a medical device not to be a device for “cleaning surfaces” (a result of the claim interpretation). 8. The Court states, with respect to inventive step: “As a general rule, a claimed solution must be considered obvious when the skilled person would take the next step prompted by a pointer in the realistic starting point or as a matter of routine to arrive at the claimed invention” . 9. The Court rules that the patent is valid and infringed. It thereby disregards a report from a Chinese Laboratory as irrelevant because it does not measure the relevant claim element (pore size). 10. In the second case (EP 485), the Court did not follow the claim construction proposed by the patentee and did not find infringement. Comment 1. The Court ordered that the Objection should not contain more words than the application and that an invalidity attack should be limited to three defences. Where such an attack is made, 25% more words may be used. 2. Nevertheless, especially the first case became rather complicated about how to measure the pore size with several technical reports. However, this did not prevent the Court from reaching what I consider to be a well-reasoned decision. 3. Once again, this case shows that it is not easy to invalidate a patent in the UPC on the basis of lack of inventive step. The Court requires a convincing argument about the incentive to combine! 4. Furthermore, in both cases, the arguments about claim interpretation (the defendant’s in the first and the claimant’s in the second case) were rather unconvincing. Sometimes, it is not bad to test your case in a mock trial with a neutral knowledgeable mock judge, because I am quite sure that the parties’ arguments were not going to win the war!

Full Decision Text

1 Local Division The Hague UPC-CFI-2594/2026 UPC-CFI-2595/2026 Order of the Court of First Instance of the Unified Patent Court delivered on 27 October 2026 In case UPC-CFI-2594/2026 and in case UPC-CFI-2595/2026: APPLICANTS 1) Versuni Holding B.V. Claude Debussylaan 88, 1082 MD, Amsterdam, The Netherlands 2) Versuni Netherlands B.V. Claude Debussylaan 88, 1082 MD, Amsterdam, The Netherlands represented by A.D. de Leeuw, I. Lummer, J. Santman, P. Willey, N.S.W. Horsting (of Vossius&Brinkhof) DEFENDANTS 1) Dreame Technology Netherlands B.V. Spicalaan 29, 2131 JG, Amsterdam, The Netherlands 2) Dreame International (HongKong) Limited Room H28G, Block E, 10th Floor, Golden Bear Industrial Centre, 66-82 Chai Wan Kok Street, Tsuen Wan, Hong Kong, China represented by S. Dahm, N.Schopp, A.Kamps and S.Ullmann (of Kather Augenstein), supported by J.P. Frederiksen and N.B. Jørgensen (of Ispicos) PATENTS AT ISSUE: EP3169210 (application 2594/2026) and EP4274458 (application 2595/2026) PANEL/DECIDING JUDGES The order is delivered by the panel of the LD The Hague consisting of the presiding judge Kokke acting as judge-rapporteur, the legally qualified judge Van Peursem, the legally qualified judge Severinsson and the technically qualified judge Papa. LANGUAGE OF PROCEEDINGS: English SUBJECT-MATTER OF THE PROCEEDINGS: applications for provisional measures 2 I. BACKGROUND AND SUMMARY OF FACTS IN BOTH CASES Joint oral hearing and final order 1. This order concerns two parallel applications for provisional measures before the same panel, concerning the same parties and the same allegedly infringing products, based on two different patents. In the application pending as case UPC_CFI_2594/2026 the applicants rely on European patent EP3169210 (“EP210”). The application pending as UPC_CFI_2594/2026 is based on EP4274458 (“EP458” and together with EP210: the “patents”). The two cases shall be referred to as the EP210-case and the EP458-case. 2. In the interests of the proper administration of justice and on account of the connection between the two applications, it was decided to hear the applications together at one oral hearing (OH), with one part dedicated to (legal) issues common to both cases, another part to the technical aspects of the EP210-case and yet another part to the technical aspects of the EP458-case. This final order also concerns both cases. The parties and the allegedly infringing product 3. The applicants, collectively “Versuni”, are a spin-off of the household appliances division of Koninklijke Philips N.V., Philips Domestic Appliances Holding B.V. Versuni’s product range consists of a broad range of household appliances, such as cleaning devices including an electric mop the (Philips) “OneUp”, launched in early 2025. Versuni has a diverse patent portfolio with more than 10 European patents covering the technology that lies at the heart of the OneUp, including EP210 and EP458. Both patents aim to protect different features of a cleaning apparatus for surfaces (such as mops) which include a dirty liquid retrieval function. 4. Applicant 1, Versuni Holding, is the proprietor of Versuni’s patent portfolio. Applicant 2, Versuni Netherlands, is responsible for commercial operations and at least part of the group’s central technical, engineering and market-facing activities, including for the Philips OneUp. It has a license under Versuni Holding’s patent portfolio. 5. The defendants, collectively Dreame, manufacture various consumer goods. Defendant 2, Dreame International (HongKong), is the parent company of the Dreame group of companies. Defendant 1, Dreamy Technology Netherlands, is the European Representative and authorised importer for Europe within the Dreame group. Dreame recently launched an electrical mop referred to as “Mira Mop (10)”. This product is depicted below (images taken from the submissions), where it is visible that the Mira Mop comprises a cleaner head (detailed in the bottom picture) with a detachable member referred to as the ‘mop pad’. The flow of clean water is illustrated by blue arrows and the flow of dirty water by brown arrows. 3 Pre-trial activities 6. Versuni first became aware of the Mira Mop in Poland. It sent a warning letter on 11 May 2026, to Dreame and its Polish retailer concerning infringement of EP 4274457 (“EP457”) in Poland. 7. Dreame presented the Mira Mop at a trade fair in Sweden on 5 May 2026, where Dreame employees talked about launch plans in the Nordics, and was subsequently commercially launched in various Nordic/Baltic and Polish web shops later in May. Versuni subsequently sent a warning letter on 15 May 2026 to both Dreame defendants inter alia bringing to their attention the imminent infringement of Versuni’s patents EP457 and EP4491079 B1 (“EP079”), two other patents in its portfolio (the “other patents”), in the Netherlands, Germany, France, the UK and Turkey, and requested Dreame to confirm that it would refrain from marketing the Mira Mop in these countries where the other patents (as well as the patents relied on in the present cases) are in force (the “First Warning Letter”). 4 8. On 26 May 2026, the German law firm Kather Augenstein responded to Versuni’s Polish warning letter and stated that it would represent Dreame International (HongKong) in IP related matters. When asked whether they would also represent Dreame in the matter relating to the other territories, Mr Dahm of Kather Augenstein responded by email on 3 June 2026 that he would indeed represent Dreame, and that he needed more time (until 22 June 2026) to formulate a detailed response. On 5 June 2026, Versuni granted this extension, provided that Mr Dahm would confirm that he was authorised and willing to accept service on behalf of Dreame should litigation become necessary, and on the mutual understanding that the extension would not affect the urgency of the matter. This was confirmed on behalf of Dreame on 9 June 2026. 9. Versuni managed to make a test purchase of the Mira Mop with a Lithuanian retailer, which purchase was delivered to Germany and arrived at Vossius’ offices 11 June 2026. Versuni subsequently had that product analysed for infringement and compared it to a Polish test purchase. 10. On 22 June 2026 Dreame provided a substantive response to the First Warning Letter concerning the other patents in which it denied infringement (with very little reasoning) and stated that the other patents would lack novelty and inventive step without providing any reasoning. Dreame ignored Versuni’s request for confirmation of Dreame’s commercial intentions with the Mira Mop in the countries where the (other) patents are valid. Unlike Dreame’s response to the Polish warning letter, in which it was explicitly stated that Dreame would “continue to market and distribute the product in question”, the response to the UPC warning letter contained no such language. 11. In order to clarify Dreame’s intentions, Versuni sent a Second Warning Letter on 24 June 2026, at the same time rebutting Dreame’s non-infringement arguments. It explicitly mentioned that failure to confirm that the Mira Mop will not be marketed in these countries will be seen as an imminent threat of (further) infringement. 12. On 30 June 2026, Dreame responded, stating that they were not convinced by Versuni’s infringement arguments and maintained its position on validity (of the other patents), adding that it would be “willing to accept a free licence”. If Versuni would not agree to a free licence by 7 July 2026, Dreame announced revocation actions against the other patents. 13. With the recently acquired test purchase, Versuni had in the meantime been able to do more testing and further infringement analyses and concluded that the Mira Mop infringes even more patents. This prompted Versuni to send a Third Warning Letter to Dreame on 9 July 2026 in an attempt to avoid unnecessary proceedings. Versuni therewith informed Dreame that the Mira Mop also infringes the patents (and two further patents), and urged Dreame to reconsider its lack of confirmation not to enter the territories where the patents are valid. Considering the amount of additional patents relevant for infringement, Versuni gave Dreame a response time until 22 July to confirm a cease and desist for all patents. 14. Dreame responded on 14 July 2026 that it would not meet the deadline of 22 July, but it would take an additional 15 days to respond, i.e. until 6 August 2026. 15. Versuni then filed the present applications for provisional measures on 17 July 2026. 5 II. PROVISIONAL MEASURES SOUGHT, SUBMISSIONS OF THE PARTIES AND PROCEDURE IN BOTH CASES Procedure 16. The Applications were filed on 17 July 2026. Subsequently, Dreame were invited to file an Objection to the Application pursuant to R. 209.1(a) on or before 29 August 2026, with the instructions that (i) the material part of the Objection cannot contain more words than the Application and (ii) in case an invalidity defence is raised, this should be concise and limited to a maximum of three attacks; in the case of such defence, 25% can be added to the number of words. 17. After the objections were filed (and the documents were adjusted to the permitted word-count), Versuni was provided the opportunity to file a further submission in reply on 10 September 2026, which Versuni did. Dreame filed a ‘rejoinder’ in both cases on 22 September 2026. 18. Dreame’s R. 109 application, timely filed of 28 August 2026, requesting simultaneous interpretation at the Court’s expense or at the expense of the proceedings, was rejected. In accordance with R. 109.4, Dreame were permitted to engage an interpreter at their own expense. 19. The oral hearing in both cases was held on 29 September 2026. Orders sought and submissions 20. In both the EP210-case and the EP458-case Versuni, asserting that Dreame infringe the patents with the Mira Mop, requests that the Court, by directly enforceable decision (and after limitation of the requests at the OH): A. order an injunction against Dreame to refrain from infringing [the patent at issue] in any way, including by making, offering, placing on the market and/or using and/or importing or storing for those purposes, the Mira Mop in the territories of the Netherlands, Germany and France (art. 62(1) UPCA and R. 211(1)(a) RoP); B. order Dreame, within two weeks after service of the decision to be rendered in these proceedings, to provide to Versuni (p/a Brinkhof N.V.) a written account with information on: (1) the origin and distribution channels of the Mira Mop, or any other product that falls within the scope of protection of EP 3 169 210 within the territories of the Netherlands, Germany and France (Art. 67(1)a UPCA); (2) the identity, including full names and address details, of distributors, resellers, retailers and other legal entities involved in the sale, supply or distribution of the Mira Mop, or any other product that falls within the scope of protection of EP 3 169 210 within the territories of the Netherlands, Germany and France (Art. 67(1)c UPCA); C. order Dreame to pay Versuni’ costs, within two weeks of the serving of the Order via CMS, of € 225.000 EUR, as agreed by the parties D. To order Dreame to pay the Court a penalty payment of € 50,000,-- for each day or part of a day that one of the granted aforementioned injunctions under A. are not complied with, and a penalty payment of € 50,000,--. 21. In the EP210-case Versuni bases its requests on infringement of claims 1-5, 7 and/or 9 of EP210. In the EP458-case it asserts infringement of claims 1, 2, 7, 10, 13, 14 and/or 15 of EP458. All requests are territorially limited to the UPCA contracting member states where the patent is valid: the Netherlands, Germany and France (the “Relevant Territory”). 6 22. In both cases Dreame requests: I. that the Application for provisional measures be dismissed; II.in the event that the Court orders interim measures, that the Applicants be required to provide security for the enforcement of a preliminary injunction and/or other interim measures "in an amount to be determined by the Court, provided that the security shall amount to EUR 5,000,000"; III. in the event that Application for provisional measures is dismissed or withdrawn, that the Applicants be ordered to bear the costs of the proceedings as agreed between the parties. 23. Dreame argue that: • Versuni Netherlands have no standing to sue. • The application lacks urgency and necessity. • The Mira Mop does not implement the claimed inventions and does not infringe the patents. • The validity of the patent is questionable because: - in the EP210-case: claim 1 is anticipated by CA2141855C (CA855), is obvious starting from FR2299001A1 (FR001) and JP2002-000530A (JP530) combined with common general knowledge or several other prior art documents - in the EP458-case, the patent is anticipated by E1 (EP1153566A2), E2 (EP3366182A1) and/or E3(JP2002000530A). III. GROUNDS FOR THE ORDERS III.1 IN BOTH CASES III.A - SUMMARY AND POINTS AT ISSUE 24. The applicable legal standard for granting provisional measures is set out in R. 211.2 RoP: the Court must be satisfied ‘with a sufficient degree of certainty that the applicant is entitled to commence proceedings pursuant to Article 47 UPCA, that the patent in question is valid and that his right is being infringed, or that such infringement is imminent’. The Court of Appeal of the UPC (“CoA”) clarified1 that ‘a sufficient degree of certainty’ pursuant to R. 211.2 RoP, in conjunction with Art. 62(4) UPCA (see also Art. 9(3) Directive 2004/48/EC) requires that the court considers it on the balance of probabilities at least more likely than not that the applicant is entitled to initiate proceedings and that the patent is infringed. A sufficient degree of certainty is lacking if the court considers it on the balance of probabilities to be more likely than not that the patent is not valid. 25. The burden of presentation and proof for facts allegedly establishing the entitlement to initiate proceedings and the infringement or imminent infringement of the patent, as well as for all other circumstances allegedly supporting the applicant's request, lies with the applicant, whereas, unless the subject-matter of the decision is the ordering of measures without hearing the defendant pursuant to Art. 60(5) in conjunction with Art. 62(5) UPCA, the burden of presentation and proof for facts concerning the lack of validity of the patent and other circumstances allegedly supporting Dreame's position lies with Dreame. 26. In the cases at hand, the Court is convinced with the required degree of certainty that the applicants are entitled to commence proceedings pursuant to Art. 47. In the EP210-case, the 1 For the first time in order of 26 February 2024 in case UPC_CoA_355/2023, NanoString v 10X Genomics. 7 Court is convinced with the same degree of certainty that EP210 is valid and that the patent right is being infringed, or that such infringement is imminent. In the parallel EP458-case, the Court’s provisionally finds that the applicable threshold for infringement with the Mira Mop has not been met. The reasons for this are explained below. In this part III.1, issues common to both cases are addressed. (Technical) Issues of the EP210-case are discussed in part III.2 below and the court sets out (technical) aspects relevant only for the EP458-case in part III.3. III.B – URGENCY, NECESSITY AND OTHER PRELIMINARY ISSUES Preliminary issues 27. The applications are admissible as the requirements set out in R. 206.2(b) to (e) RoP are complied with. 28. Defendant 1 resides within UPCA-territory and Dreame are asserted to (imminently) infringe European patents in, inter alia, the Netherlands. International jurisdiction and competence were not challenged. Dreame are thus considered to submit to the jurisdiction of the UPC and to the competence of the LD The Hague. Ex officio, the Court notes that international jurisdiction can be based on art. 26 BR2 for both defendants. 29. After Dreame initially raised objections regarding Versuni’s standing to sue, these were not maintained after Versuni’s subsequent clarifications. The Court is convinced with a sufficient degree of certainty that both applicants are entitled to commence proceedings pursuant to Art. 47. Unreasonable delay? 30. Dreame asserts that the applications should be dismissed because they were filed with unreasonable delay after Versuni became aware of the Mira Mop at a trade fair in Sweden on 5 May 2026 and in Poland, and in any case after it sent the First Warning Letter on 15 May 2026. 31. According to R. 211.4 RoP, the Court shall have regard to any unreasonable delay in seeking provisional measures. This delay shall be calculated from the day on which the applicant became aware, or should have become aware, of the infringement in a way that would enable it, in accordance with R. 206.2 RoP, to file an Application for provisional measures with a reasonable prospect of success. Thus, the decisive point in time is when the applicant has, or should have had, after exercising due diligence, the reasonably necessary facts and evidence within the meaning of R. 206.2(d) RoP.3 32. It follows from the above that an applicant does not need to apply to the Court until it has – or should have had – reliable knowledge of all the facts which make an action for interim measures likely to succeed including evidence to credibly substantiate those facts. The applicant may prepare for any possible procedural situation that may arise in the circumstances in such a way as to be able to submit the requested information and documents to the Court upon an appropriate order and to successfully respond to the opposing party’s submissions. On the other hand, the applicant must not delay unnecessarily. As soon as it becomes aware of the alleged 2 RegulaJon (EU) no 1215/2012 of the European parliament and of the council of 12 December 2012 on jurisdicJon and the recogniJon and enforcement of judgments in civil and commercial maQers (recast). 3 CoA, 25 September 2024, UPC_CoA_182/2024, Mammut vs Ortovox confirmed in inter alia CoA 13 August 2025, UPC_COA_446 and _520/2025, Boehringer Ingelheim vs Zentiva, CoA, 30 March 2026, UPC_CoA_899/2025, Sinocare v Abbott and CoA, 2 July 2026, UPC_CoA_19/2026, Guardant v Sophia. 8 infringement, it must investigate it, take the necessary steps to clarify the matter and obtain the necessary documents to support its submission. In doing so, the applicant must pursue the necessary steps with determination and bring them to a conclusion. Once the applicant is in possession of all the knowledge and documents that are reasonably likely to lead to a successful prosecution of the case, it must normally file an application for provisional measures without delay.4 33. In this case, Versuni acted prudently by basing its action before the UPC not on a Mira Mop obtained outside UPCA territory in early May 2026, but on a test purchase delivered to Germany, within the Relevant Territory. Versuni obtained this test purchase with sufficient diligence on 11 June 2026 (as described above at 9), at the same time trying to obtain a confirmation from Dreame that they would not launch the Mira Mop in the Relevant Territory. Dreame did not provide such confirmation. Versuni in the meantime subjected the Mira Mop delivered to Germany to a thorough investigation, also on the inside, whereby infringement of the patents at issue was considered positively ascertained, in addition to the two patents originally asserted based on the Mira Mop acquired earlier and relied on in the First and Second Warning Letters (see above). After this information was obtained, the Third Warning Letter was sent on 9 July 2026 mentioned the patents for the first time, and the applications were filed with undue delay on 17 July 2026. Dreame’s urgency-defence it thus dismissed. III.2 IN THE EP210-CASE (UPC-CFI-2594/2026) III.C THE PATENT, BACKGROUND AND TEACHING 34. EP210 was granted on 16 May 2018 for a “Surface cleaning device”. The application for EP210 was filed on 8 July 2015 (first published as WO2016008773A1), claiming priority of 14 July 2014. EP210 proceeded to grant with the same text and drawings as originally filed: no amendments to either the claims or description/drawings were introduced during examination. No opposition was filed. 35. EP210 relates to the field of surface cleaning devices (para. [0001]). The introductory part of EP210 discusses the disadvantages of prior art manual cleaning devices, paying particular attention to the step of picking up the dirt including part of the liquid initially supplied to remove said dirt (paras. [0006]-[0007]). In particular, it is noted that: • the main disadvantage of the bucket-and-mop system is that the amount of liquid that remains on the floor is difficult to control, as it depends upon the force/action exerted by the user when wringing the mop and applying it on the floor; in addition, a mop which is too dry can result in a poor cleaning performance, while a mop which is too wet may cause damages to the floor (para. [0008]); • the main disadvantage of window cleaning with sponge and squeegee is that the liquid which is wiped from the window is not collected and runs down the window glass - or the user needs to collect it manually by a sponge or cloth (para. [0009]); • a general disadvantage of the above-mentioned manual cleaning devices is that their performance over time is not constant, e.g. due to the mop becoming increasingly dry or wet during a cleaning cycle (para. [0010]). 4 See e.g. LD Munich 21 May 2024, UPC_CFI_443/2023 and 27 August 2024, UPC_CFI_201/2024. 9 36. The introductory part of EP210 also discusses the electric cleaning appliances, which overcome the above shortcomings of the manual devices, in that they use suction/airflow to wipe the liquid from the surface to be cleaned and transport such liquid to a reservoir. However, according to EP210 said electric systems also have disadvantages, namely they (para. [0011]): • need a rather big fan/motor to generate a sufficient airflow to draw dirt liquid from the surface to be cleaned; • require the air and the liquid to be separated again once the liquid has been collected from the floor; • require the fan/motor being suited to handle moist or moist air; • in general, also in view of the above points, are expensive and complex. 37. The object of the invention is stated as providing an improved surface cleaning device. The invention is defined by the independent claim, with advantageous embodiments defined in the dependent claims (para. [0012]). 38. According to EP210, the disadvantages of the prior art are solved by a device having “a new and innovative way of liquid transport” (para [0014], first sentence). In particular, such “liquid transport” operates as follows: a cloth absorbs the liquid from the surface to be cleaned; the cloth is placed directly onto a porous material, which has pores of a specifically selected size and which is in fluid communication with, or part of, a reservoir; a dedicated arrangement generates an under-pressure in the reservoir to draw liquid therein. This configuration is explained at para. [0014], partly copied below (emphasis added). [0014] […] The amount of liquid the cloth can absorb is rather limited but therefore this cloth is placed directly on a porous material. The porous material is part of a reservoir which can contain liquid. A small under-pressure is applied to the reservoir. When the cloth is wetted due to picking up liquid from the cleaned surface, a certain amount of liquid is sucked out of the cloth and will pass through the porous material into the reservoir due to the pressure difference. This cloth is therefore continuous kept dry/moist. Because the amount of liquid which can be absorbed is limited by the reservoir capacity rather than by the amount of liquid that can be absorbed by the cloth itself, an appliance can be built with the simplicity of a squeegee with a sponge without the difficulties of a double squeegee nozzle, so that no large fan or complex separation method is needed. An additional advantage of the above system is presented in para. [0015]: [0015] Another advantage of this embodiment is that the cloth can be sucked almost completely dry by applying more under-pressure. This means that the appliance can be stored with an almost dry cloth, which is much more hygienic than a wet cloth. 39. The reduced size and complexity of the under-pressure applying system mentioned at the end of para. [0014] can also be better understood with reference to para. [0020], according to which “the pore size is chosen in such a manner that the material has a preference to pass liquid above air when a pressure difference is applied”. 40. The core of the proposed solution can be summarised as follows: - because the cloth is placed (directly) on a porous material with a specific (claimed) pore size, when the cloth becomes wet a controlled amount of liquid – but none or only little air – passes from the cloth through the porous material and into the reservoir; the properties of the porous material allow obtaining such liquid drawing into the reservoir by applying a modest under-pressure (para. [0014]); 10 - in this way, the cloth can be kept dry or moist in a controlled way (paras. [0014], [0016]); - as said above, the arrangement of a cloth in contact with a porous material of selected pore size allows liquid (not air) to be primarily transported, so that no liquid- air separation is needed; therefore the appliance can be of a simple, energy-efficient and cost-effective construction (para. [0020]); - the combination of two layers with different pore structures – namely the cloth and the porous material – allows wiping the surface and collecting liquid by the cloth and then transporting the liquid via the porous material to a reservoir without it being allowed to pass back into the cloth due to the choice of pore size of the porous material (paras. [0022], [0023]; claim 5). 41. In short, EP210 teaches that the interposition of a porous material of selected pore size between the absorbing cloth and the sucking system allows maintaining a controlled pressure difference between the reservoir and the cloth, so as to withdraw a controlled amount of dirt liquid from the surface being cleaned. The porous material may also impede dirt liquid to flow back and wet the cloth. 42. EP210 has one independent claim 1, and dependent claims 2-9, directed at a surface cleaning device. Claim 1 is divided into features as follows: 1.1 Surface cleaning device, comprising: 1.2 a cloth (C) placed on a porous material (PM), 1.3 a pore size of the porous material (PM) being between 1 μm and 50 μm; 1.4 a reservoir (R) for collecting liquid absorbed by the cloth (C); and 1.5 an arrangement (P) for applying under-pressure in the reservoir (R) so as to transfer liquid from the cloth (C) into the reservoir (R). 43. Fig. 1, copied below, shows the basic structure and operating principle of the claimed device. The cloth is denoted by C and is disclosed as in direct contact with the porous material PM (paras. [0020], [0024]). A reservoir R covers the whole porous area (para. [0025]). An arrangement P provides under-pressure over the surface formed by the porous material PM, which is in contact with the wet cloth C, in order to suck moisture that is captured in the cloth C while cleaning. The arrangement P can be a simple syringe or an electrically powered electric pump. (Para. [0026]). 11 44. Figs. 2 and 3, depicted below, provide specific examples of hand-held, window cleaning devices including the claimed subject-matter (as indicated also at point 40 of the APM). In the embodiment of Fig. 2 (shown on the left), a squeegee S collects dirt liquid L from a window, which liquid is absorbed by the cloth C, and, via the porous material PM, sent in the reservoir R due to under-pressure applied by the pump P. The latter is an electrical pump, powered by batteries B. A switch SW serves to switch the pump P on/off (cf. [0029]). In the embodiment of Fig. 3, the dirt liquid is collected in the body of the window cleaning device, thus implementing a larger reservoir R. A separate tank can also be provided (cf. [0030]). III.D CLAIM CONSTRUCTION, SKILLED PERSON 45. The interpretation of the patent claims and their features is a matter of law and is based on Art. 69 EPC and the Protocol on its Interpretation. The patent claim is not only the starting point, but the decisive basis for determining the protective scope of the patent. The interpretation does not only depend upon the strict, literal meaning of the claim wording. Rather, the description 12 and drawings must always be used as explanatory aids for the interpretation of the patent claim and not only to resolve any ambiguities therein. However, this does not mean that the patent claim merely serves as a guideline and that its subject-matter also extends to what, after examination of the description and drawings, appears to be the subject-matter for which the patent proprietor seeks protection. The patent claim is to be interpreted from the point of view of a person skilled in the art. In applying these principles, the aim is to combine adequate protection for the patent proprietor with sufficient legal certainty for third parties. These principles for the interpretation of a patent claim apply equally to the assessment of the infringement and the validity of a European patent.5 The skilled person always interprets the features of a claim in the light of the claims as a whole.6 46. The imaginary skilled person in this case is a mechanical engineer at BSc level, with approximately 5 years of experience in the development of wet cleaning apparatus, and who is able to consult a team providing relevant competences from other fields as required by the circumstances, such as the fields of porous materials, battery-powered electrical systems, and pumps. Even though the outcome does not depend thereon, the Court adopts the definition proposed by Dreame, which is in line with, but slightly more detailed than, the definition proposed by Versuni and includes knowledge of related fields of which the skilled person must be expected to be aware.7 47. The parties disagree on the interpretation of several features. Insofar as these are relevant for this order, these are discussed below. Feature 1.1 – “Surface cleaning device” 48. Dreame asserts that the “cleaning device” of claim 1 can be any liquid removal device, which can also include, e.g., medical absorbing arrangements (as disclosed in prior art document CA855). Versuni disputes this interpretation, noting that the term “cleaning”, also as read in the context of the patent as a whole, limits the claimed invention to household appliances. It also points out that the functions of the surface cleaning device are not limited to liquid removal. 49. Dreame’s interpretation is not followed. The overall patent disclosure – including the described embodiments and stated technical effects / advantages over the prior art – is clearly directed to devices for cleaning surfaces in an external environment, e.g. household appliances. The skilled person will derive this inter alia from (paras. [0016] and [0032]), where floors and windows as examples of surfaces to be cleaned are mentioned, indicating that any surface orientation or shape (“flat or bumpy”) can be handled. 50. Further confirmation is found in para. [0006] (copied below) which can be considered a definition for the claimed “surface cleaning” application. 5 CoA orders of 26 February 2024 and 11 March 2024, UPC_CoA_355/2023, 10x Genomics v NanoString and CoA order of 30 April 2025, UPC_CoA_768/2024, Insulet v EOFlow. 6 CoA order of 13 May 2024, UPC_CoA_1/2024, VusionGroup v Hanshow; CoA order of 30 April 2025, UPC_CoA_768/2024, Insulet v EOFlow and CoA order of 25 November 2025, UPC_CoA_464/2024 (and other numbers), Meril v Edwards. 7 CoA order of 25 November 2025, UPC_CoA_464/2024 (and other numbers), Meril v Edwards, headnote 10. 13 [0006] For cleaning surfaces in general, three steps can be identified. Step 1: bringing a liquid to the surface. Step 2: an agitation action for loosening and dissolving the dirt. Step 3: taking up the dirt including part of the liquid. For this last step, there are in general two solutions. […] Therefore, the patent specification does not support the interpretation of feature 1.1 as including the absorption of body fluids, e.g. from the skin, by a personal care device. The term 'surface' would not be understood to encompass this by a skilled person. 51. The EP210 specification also does not support limiting the functions of the cleaning device to liquid removal only, the final step mentioned in [0006]. The ‘cleaning’ denotation of the claimed device entails the device including other features that make it suitable for performing the cleaning of surfaces, in addition to removing (dirty) liquid. Even if not all the claimed embodiments explicitly provide for an agitation action or fluid delivery according to the definition of para. [0006], and the cleaning functions are not explicated in the claim, the term ‘surface cleaning’ is clear in itself. The skilled person would therefore interpret feature 1.1 as referring to household cleaning devices suitable for cleaning surfaces such as floors or windows, thereby removing dirty liquid. Feature 1.2 – a cloth (C) placed on a porous material (PM) feature 1.2 52. This feature requires that the cloth is “placed on” the porous material. Dreame asserts that direct contact between the cloth and the porous material is necessary and that such contact must extend across their entire area in order to support the under-pressure and to ensure that very little or no air (para. [0026]) is permitted to enter the reservoir. 53. The wording of the claim requires neither a direct contact nor a connection or superimposition of the cloth and the porous material over their entire area. It is correct that in all the disclosed embodiments the cloth and the porous material are presented as in direct contact, i.e. with no interposition of other layers between them (see, e.g., para [0014]: ‘[…] this cloth is placed directly on a porous material […]’; para. [0016]: ‘A cloth is placed in direct contact of the porous material’; para. [0024]: “The cloth C that is intended for the cleaning/drying action is in direct contact with the porous material PM.”). However, such (direct) contact is not claimed. Moreover, the technical effect associated with the claimed arrangement, i.e. controlling the dirty liquid withdrawal, is compatible with the optional use of intermediate layers, with appropriate permeability properties, between the cloth and the porous material, as well as with only a partial overlying of the two materials. In addition, the connection modes disclosed in para. [0017] of EP210 recited above do not (all) imply a continuous connection (see, e.g., the recitation of a “Velcro strip”). Finally, Versuni correctly notes that an absolute barrier to air is not required by the patent disclosure. Feature 1.3 - a pore size of the porous material (PM) being between 1 μm and 50 μm 54. In dispute in the context of this feature is inter alia what the technical effect thereof is. The effect that the material with this pore size-range has a preference to pass liquid above air when a pressure difference is applied, is not in dispute. But Dreame asserts that the porous material 14 claimed also functions as a (one-way) valve, avoiding a reverse liquid flow, with reference to the following paragraphs in the description of EP210: [0022] […] The porous plastic is important as it has two functions: a. Allow moisture from the cloth C to pass to the reservoir R because of under-pressure; b. Do not allow the moisture collected in the reservoir R to come out again, i.e. allow moisture in one direction only and prevent from expelling it in the opposite direction. [0023] This aspect of not expelling the moisture in another direction is indeed very important as otherwise the cloth C will become wet again. So if we remove the porous plastic and use a cloth or sponge alone, the under-pressure will suck liquid but will wet the cloth again. Also, the porous plastic will help to create a better pressure difference which cannot be achieved using the more porous/open structure of a sponge or cloth. 55. Dreame’s position that feature 1.3 implies that the porous material is designed to function as a (one-way) valve is dismissed. The skilled person will appreciate that this disclosure refers to a specific implementation of porous plastics as the porous material. The skilled person will also note that a specific embodiment in which the porous material does not allow the collected liquid to come back through the pores is claimed separately in dependent claim 6. Therefore, the capability of (fully) avoiding a two-directional flow is not necessarily implied by the pore size selection according to claim 1. 56. Whether the claimed pore size-range refers to minimum, maximum or mean pore sizes, is also in dispute. This interpretation issue does not need to be decided. This could be relevant for infringement but, as Versuni rightly pointed out during the OH, the tests relied on (to be discussed below) show that the attacked product falls within the claimed range regardless whether the range refers to minimum, maximum or mean Feature 1.4 – a reservoir (R) for collecting liquid absorbed by the cloth (C) 57. There are two issues of interpretation in dispute relating to the claimed ‘reservoir’ of feature 1.4, namely: (i) whether, functionally, such a reservoir is intended to store a significant amount of liquid for a certain period of time, or whether a simple intermediate collector downstream of the porous material, from which liquid is continuously withdrawn by suction, qualifies as a ‘reservoir’; and (ii) structurally, regarding the relative positioning of such a reservoir with respect to the porous material, in particular whether the reservoir should include the entire area downstream from the porous material. 58. The skilled person understands that the reservoir of feature 1.4, interpreted in the context of the claim as a whole (especially also feature 1.5) and of the entire patent specification, is functionally defined as a space suitable for collecting liquid retrieved from the cloth using under- pressure. This follows inter-alia from the requirement that under-pressure is applied to the reservoir in order to transfer liquid from the cloth into the reservoir (feature 1.5). Therefore, any space suitable for collecting the retrieved liquid and to which under-pressure is applied qualifies as a reservoir according to this feature. There do not appear to be technical reasons – nor have these been presented by Dreame – that require a strict proximity of the complete storage to the porous material or that exclude a small reservoir in proximity to the porous material and a larger 15 tank located remotely. For the same reasons, it is not required that the ‘reservoir’ holds the liquid for a certain time. In other words, also a collector along a transport line may function as the claimed ‘reservoir’. 59. This follows from the patent specification. The skilled person learns that the porous material may either be a part of the reservoir (as claimed in dependent claim 5), or the porous material can be separate from the reservoir (e.g. in para [0022]: ‘The porous material PM is a part of the reservoir R or part of a soleplate that is a housing for the cloth or sponge to rest upon’; see also Fig 1 and Fig. 3; paras [0016], [0025] and [0030]) and as described, e.g., at par. [0022] in conjunction with Fig. 1: ‘The porous material PM is a part of the reservoir R […]’. In several disclosed embodiments, the porous material is a portion of the reservoir but does not per se cover the full reservoir. Alternatives are described in para. [0025] ‘[…] In another embodiment, a smaller housing may house the porous material PM and the cloth C, and then channel (e.g. through gravity) the liquid passing through the porous plastic to a reservoir formed by a container downstream for holding dirty liquid. In this case, the capacity to hold liquid may be larger.’ and in para [0030] in conjunction with the embodiment of Fig. 3: ‘Fig. 3 shows another embodiment of the window cleaning device in accordance with the present invention wherein the waste liquid is not contained directly behind the porous material PM assembly. In this embodiment, the waste liquid is contained in the body of the window cleaning device as compared to the device shown in Fig. 1. This means that the device will have a larger reservoir R or a separate tank where the waste liquid is collected. The advantage offered by these configurations is that more liquid can be collected and stored allowing a fairly longer operation time.’ The above passages clarify that the reservoir can include a separate container not necessarily located in the immediate proximity of the porous material and that also a storage tank separate from, and additional to, the reservoir can be provided, in order to increase the liquid storage capacity, as long as (part of) the reservoir is suitable for collecting waste liquid using under- pressure. III.E – VALIDITY OF EP210 Novelty 60. Dreame asserts that claim 1 is fully anticipated by the prior art document CA2141855C (CA855; Exhibit KAP7). It relies in particular on the embodiment of claim 30 of CA855 in combination with Fig.2a. Claim 30 of CA855 reads as follows: A liquid removal system, comprising: an interface device, said interface device comprising an entrance zone and a second zone and having a porous membrane separating said entrance zone from said second zone, said second zone having a vacuum source connected thereto by means of a drain tube; said porous membrane being capable of maintaining a vacuum in said second zone without permitting air from said entrance zone to pass through said membrane into said second zone when said membrane has been wetted with a first liquid; said vacuum being maintained until said membrane is contacted with a second liquid; and wherein said second liquid upon entering said entrance zone and contacting said porous membrane is removed from said entrance zone by said vacuum in said second zone by passing through said porous membrane into said second zone. 16 Fig 2A, annotated by Dreame, is copied below: 61. CA855 discloses a liquid removal system for use in “laboratory, workshops and elsewhere” (see, e.g., page 1, first and last paragraph). The disclosure is directed to the medical field including personal care items, e.g. catheters (page 1, second paragraph). The object of the disclosed system is a prompt and effective removal of (body) liquids (esp. urine) by avoiding contact with the skin or personnel. 62. Versuni rightly argued that CA855 does not disclose the ‘surface cleaning device’ of feature 1.1 of claim 1, when construed correctly as discussed above. CA855 does not pertain to household appliances. In addition, the only function of the liquid removal system of CA855 is to remove liquid. The (first two) cleaning functionalities that the skilled person will understand to be an inherent part of the claimed surface cleaning device, are not – explicitly or implicitly – disclosed in the system of CA855. It is thus considered more likely than not that this novelty validity attack will fail. Inventive step 63. For inventive step, Dreame initially relied on two prior art disclosures as starting point, namely FR2299001A1 (FR001; Exhibits KAP8, KAP8a) and JP2002-000530A (JP530; Exhibits KAP9, KAP9a) in different combinations. When reminded at the oral hearing that each combination is considered a separate attack, and only three attacks in total were permitted in the context of these proceedings regarding provisional measures, Dreame indicated to rely on FR001 and combinations only, in addition to the novelty attack. 64. For the assessment of inventive step, considering the discussion of the prior art contained in EP210 (paras. [0008]-[0011]) and the stated solution (paras. [0012], [0014], [0015]) as summarised in 40. above, and taking into account the test developed by the CoA of the UPC8, the skilled person will understand that the objective problem/technical teaching of the invention is to provide an improved surface cleaning device which is suitable to overcome the drawbacks of the known systems in terms of controllability of the amount of dirt liquid that is drawn from the floor – and eventually wets and is retained in the cloth – and which at the same time does not require large and complex motors or fans. This differs slightly from the objective problems suggested by the parties, but these formulations do not change the assessment of inventive step. 65. FR001 discloses a surface cleaning device capable of dispensing a cleaning liquid and collecting dirty liquid. The object of FR001 is that of providing a lightweight and easy-to-use mopping device capable of wiping cleaned surface after washing by avoiding the use of large volumes of suction air (page 2, first paragraph of FR001a). It is considered a realistic starting point, as it 8 CoA orders of 25 November 2025 in UPC_CoA_464/2024 (and other numbers), Meril v Edwards and in UPC_CoA_528 and 529/2024, Amgen v Sanofi. 17 relates to the same type of cleaning device as EP210 and is concerned with a similar problem. Figs. 1 to 3 of FR001 are copied below as annotated by Dreame. 66. The dirty liquid is sucked through the membrane and is adducted into the bucket (35) through a pipe (21). According to FR001 – and similarly to the teaching of EP210 – the membrane is chosen with a sufficient fine porosity so that, once wet, the dirt liquid can be sucked through it without air. In this way, relatively low power means are sufficient to quickly and completely remove the liquid from the treated surface. (Page 2, fourth paragraph). 67. It is not in dispute between the parties that in any case the specific pore size of feature 1.3 (‘a pore size of the porous material (PM) being between 1 μm and 50 μm’, left out by Dreame in the claim chart comparing FR001 with the patent in its Objection), is not disclosed in FR001. 68. Dreame assert that the claimed porosity according to feature 1.3 amounts to an arbitrary selection which cannot generally contribute to inventive step. The Court understands this to mean that the skilled person would arrive at the claimed invention starting from FR001 using his common general knowledge in an obvious way. In support of the arbitrariness of the claimed range, Dreame note that, despite para. [0020] of EP210, cited below, tests showing that the claimed range is suitable to obtain the desired effect are not disclosed in EP210. [0020] The essence of the embodiment of Fig. 1 is the porous material PM and the contact area with the cloth C. The porous material PM is preferably hydrophilic, and the pore size is chosen in such a manner that the material has a preference to pass liquid above air when a pressure difference is applied. Tests showed that a pore size between 1 mm and 50 mm is suited for this application. 69. Dreame also refer to the subsequent patent publication EP458 – the subject of parallel proceedings UPC_CFI_2595/2026 – wherein several tests concerning the properties of the pores 18 are discussed and a preferred range of ‘limiting’ (i.e. maximal) pore size is indicated of 15-105 μm (paras. [0061], [0063], [0324] of EP458). As the pore size range demonstrated to achieve the stated objectives in EP458 is different from the range claimed in EP210, the choice in EP210 is arbitrary, according to Dreame. 70. Dreame notes that the claimed pore size is not “extraordinary”. This is known from other prior art documents, e.g. WO00/00281A1 (WO281; Exhibit KAP11) (claims 1 and 2, where an average pore size of 10-100 micrometers is mentioned) and CA855, which can be combined with FR001 to lead to the invention without inventive step. 71. Versuni in response inter alia point out that the high-volume flows (page 3, second paragraph), the long sheath (Fig. 1), and the large pump unit (Fig. 1) in FR001, all indicate larger pores than those claimed. Therefore, they consider that FR001 teaches away from the claimed pore size- range of feature 1.3 72. The Court finds Dreame’s argument that the claimed pore-size range can be disregarded as it is arbitrary, not convincing. EP210 clearly teaches that there is a technical (capillary) effect associated with the range: the pore size-range is chosen in such a manner that the material has a preference to pass liquid above air at a small under-pressure so that a low-power small pump with small battery life suffices (para. [0020] and [0031], see also 40 above). Even if the post- published EP458, wherein a range of is 15-105 μm is mentioned, could be taken into account, this reference does not support that the claimed pore size range does not provide the stated technical effects. The disclosure of EP458 is concerned with balancing a series of parameters, e.g. a relatively high capacity to keep an under-pressure with a relatively low resistance to liquid transport through the porous material (see, e.g., para. [0264]). However, EP458 does not contain any explicit or implicit teaching against a pore size of 1-50 μm. Moreover, the pore size range indicated in EP458 (15-105 μm) is not completely remote from the claimed one. The question is not whether the technical effect is also achieved outside the range claimed in EP210, only whether this effect is reached within the claimed range. This is not disputed and the technical effect can therefore be taken into account for inventive step. 73. The court therefore needs to assess whether the skilled person, starting from FR001 and trying to solve the objective problem defined above, would add the claimed pore-size range as a next step in an obvious way to arrive at the claimed invention. As a general rule, a claimed solution must be considered obvious when the skilled person would take the next step prompted by a pointer in the realistic starting point or as a matter of routine to arrive at the claimed invention. 74. Dreame does not explain thoroughly and convincingly why the skilled person, starting from FR001, would be motivated to choose the claimed pore size-range as a matter of routine, based on prior art or by combining FR001 with CA855 or WO281. The skilled person would not have found a pointer or motivation in FR001 to select the claimed pore size-range for the device of FR001. CA855 belongs to a completely different technical field than FR001, as discussed in the context of novelty, and the skilled person would not combine this with the cleaning device of FR001. WO281 – which discloses a different configuration of an absorption device from FR001 and which would therefore entail different configuration needs than FR001 – would also not be combined. And even if this combination were made, the pore size range disclosed in WO281, i.e. 19 1-100 μm, does not lead to the claimed range of 1-50 μm. Conclusion on validity 75. In view of the above, the Court does not consider it more likely than not that EP210 will be found to be invalid in proceedings on the merits. III.F -INFRINGEMENT 76. Versuni asserts literal infringement of inter alia independent claim 1 with the Mira Mop. Dreame disputes that the features 1.2, 1.3, 1.4, and 1.5 are reproduced in the Mira Mop. These features are discussed below, with reference to the pictures of parts of the Mira Mop copied below from the parties’ submissions. The first picture shows a cross-section of the underside of the detachable Mira Mop’s mop pad, illustrating that it is formed by three layers: a blue microfiber cloth, which wipes the floor, attached to a finer, water-permeable grey porous layer, which is in turn attached to a blue foam like layer. Clean water flows out through the holes in the center which is connected to a clean water tank by the blueish tubes shown below. The dirty water enters through the suction holes located in an elongated protrusions visible in the picture of the underside of the cleaner head of the Mira Mop. Dirty water flows out of these protrusions through the yellow tubes. The last picture shows an arrangement to create under-pressure in the area behind the suction holes. 20 77. In the context of features 1.2 and 1.3, Dreame acknowledges the presence of a microfiber material (grey part in the pictures above). However, Dreame’s argument that feature 1.2 is not met in the Mira Mop because there is no continuous attachment of the porous material to the cloth (as illustrated in the picture below), fails. This is not a requirement when this feature is construed correctly, as set out in 52 and 53 above. 78. According to Dreame, the Mira Mop uses a peristaltic pump to prevent the backflow of dirty liquid, rather than a porous material acting performing this function. Dreame’s argument that there is no infringement because the ‘porous material’ of the claim must act as a one-way valve and prevent backflow, which is not met in the Mira Mop, does not stand up to scrutiny for the same reason: the assumption that the porous material must have a one-way valve function, is based on an incorrect interpretation of claim 1, as discussed at 54 and 55. Reservoir 79. Versuni reads the two cavities of the longitudinal protrusions on two sides of the cleaner head as the claimed reservoir of features 1.4 and 1.5. This is shown in the picture below where the cleaner head is cut in half and the longitudinal cavities designated as ‘reservoir’ by Versuni are indicated with red arrows. An explanatory scheme of the fluid flow according to Versuni is shown in the following picture. The dirt inlets correspond to what is referred to as ‘suction holes’ in one of the pictures shown in 76 above. The dirty liquid entering from the five inlets on each side is, according to Versuni, sucked into the cavities by the under-pressure maintained there and consolidated/collected in these cavities before the water is transported to a larger unit for dirty water via the tube. 80. According to Dreame, the Mira Mop does not have a reservoir (R) for collecting liquid absorbed by the cloth as required by the claim. The cavities do not “collect” the liquid, but are a passageway, manifold or transit structure, towards the remote storage tank of the attacked device. In other words, still according to Dreame, the cavities should be regarded as part of the transport system and not as a reservoir (which requires storage). Rather, only the remote storage tank (the dirty water tank) further upstream in the Mira Mop can qualify as a reservoir, but no under-pressure is applied to this tank. A scheme of the attacked device with arrows 21 showing the liquid flow according to Dreame is copied below. Dreame also compares the volume of the cavities with that of the dirt and clean water tanks of the attacked device, to further emphasize the lack of any liquid “collection” in said cavities. 81. As explained above at claim construction, the ‘reservoir’ of EP210 does not require a certain storage size or a specific positioning with respect to the porous material, provided under- pressure can be applied for drawing liquid through the porous material. Nor does the claimed reservoir appear incompatible with being part of a liquid transport system within which it acts as a collector. Therefore feature 1.4 is implemented in the Mira Mop by the cavities indicated by Versuni. In the Mira Mop an under-pressure is applied in cavities through the connection tubes. The attacked device thus operates in the same way as provided by feature 1.5, which is also met. That no under-pressure is applied to the dirty water tank further downstream is irrelevant as this (additional) reservoir is not a reservoir (R) of claim 1. a pore size of the porous material (PM) being between 1 μm and 50 μm 82. Whether the porous material of the Mira Mop meets the requirement of a pore-size within the claimed range, is also in dispute. This topic is considered the most contentious issue in the debate between the parties and is crucial to the assessment of infringement. The debate is summarised below. - The parties’ positions 83. Versuni substantiates that this feature is met with a technical test report prepared by an independent third-party laboratory assessing the pore size of the attacked product (Exhibit VB-A23, the First Report). The tests were performed with the machine “Porolux 1000” in line with the industry standard ASTM F316-03(2019), using Capillary Flow Porometry (CFP) as the measurement technique. Accordingly, a wetting liquid fills the pores; an inert gas is then introduced to displace the liquid; the pressure required to empty a pore is inversely proportional to its size. The result table for maximum, mean and minimum pore size is copied below, when using Porefil or water as the wetting agent. For both liquids, the results show that the maximum, 22 mean and minimum pore size measures all fall within the claimed range. The graphs showing some flow vs diameter, as presented in the report for each liquid used, are also copied. 84. The conclusion of the Report contains the following: 4.1 Bubble point measurements with POROLUX™ 1000 The measurements with Porefil yielded an average maximum pore size of 25,0 μm, an average mean 23 pore size of 11,4 μm, and an average minimum pore size of 5,6 μm. The measurements with pure water yielded an average maximum pore size of 45,9 μm, an average mean pore size of 24,1 μm, and an average minimum pore size of 18,8 μm. 85. Dreame contests the measurements as “highly inaccurate” and not sufficient to satisfy Versuni’s burden of proof. In particular Dreame asserts that: (a) the measuring technique used, namely Capillary Flow Porometry, typically yields smooth, continuous curves mapping the test results (flow vs estimated pore diameter) when executed with adequate precision, whereas the graphs of the report (copied above) show pronounced jagged edges and spikes; this jaggedness is, according to Dreame, the direct visual evidence of a severely insufficient number of data points and of an inappropriately low measurement resolution; moreover, according to Dreame the data points are concentrated at smaller pore sizes, i.e. there is an asymmetric sampling influencing the calculation of the average size, which is artificially skewed downwards. Moreover, Dreame notes that said graphs do not reproduce the full range of measures reported in the tables copied above (namely, the plotted measurements with Porefil terminate at 9 μm). 86. In addition, Dreame asserts that the test standard (ASTM F316-03) was not applied correctly, mainly because: (b) it only covers maximum pore sizes in a range 0.1-15.0 μm, which is a narrow sub-range of the claimed one, whereas the measures provided in the Report exceed said test scope; (c) the test method is only applicable for non-fibrous filter membranes with circular cross section, whereas the porous material of the attacked product is woven from warp and weft yarns; (d) the pore size is an inherent property of the material and should not change depending on the wetting liquid used in the test, whereas the Report indicates important variations (see the tables copied above); moreover, Porefil is not a reagent indicated in said standard; as per water, the standard indicates a certain purity grade that the Report has not specified; (e) the standard indicates specific dimensions for the specimen to be tested, which dimensions are not disclosed in the Report; moreover, the porosity of the material is influenced by the presence of distracting factors like, e.g., glue, and it is not clear how Versuni could isolate a specimen of suitable dimensions without such factors from the attacked device. 87. Dreame submits their own test results conducted according to the official Chinese GB/T 6165- 2021 standard and concludes that the majority (57%) of the pores of the sample are smaller than 0.3–0.5 μm (Exhibit KAP6, the Counter Report). 88. Versuni replied to the above criticism with a second clarifying report from (Exhibit VB- A38; the Second Report), on which Second Report Dreame commented in its rejoinder. The issues were further debated during the oral hearing. The further debate between the parties can be summarised as follows. § In conjunction with the issues under above point (a), Versuni affirms that the pore size measurements in the Report are relatively low-resolution ones, adapted to understand whether the pore size falls within the claimed range of 1-50 μm; this does not mean that such measurements can be characterised as erroneous or “inaccurate”; even if the resolution would have been higher, this will not bring the pore size outside of the claimed 24 range. Dreame notes that the results are biased if the measurement scheme has been construed to specifically identify only pores in the claimed range and not to ascertain what the actual pore-size distribution is. § Still in conjunction with point (a) and concerning the number and distribution of data points displayed in the Report’s graphs, according to Versuni, Dreame has not made credible that additional plotted points would bring the product outside of the claimed range. At the OH, Versuni explained that the cut-off of the graphs to diameter values lower than those indicated in the corresponding tables is due to a partial representation of data obtained by the machine software. Dreame considers that the burden of proof remains upon Versuni and that it has not been met. § Regarding the above issues under (b) to (e), Versuni notes that the bubble point and pore size measurements were performed in accordance with best practices and were simply “based upon” the standard ASTM F316-03 (which standard was not mentioned in the first Report). In particular, the Second Report specifies that the reference to the standard was included “for the way in which the bubble point measurement and pore size were calculated in the Report”. Dreame objects that it is not clear what “based upon” means. § In conjunction specifically with point (b) (and also in further confirmation of the reply to the issues raised under (a)), Versuni asserts that the “Porolux 1000” machine allows for the measuring of pores within a size range of approximately 0.09 - 150 μm (as indicated at page 2 of the Report). At the OH, Versuni confirmed that the test according to the Report has classified all the pores in the sample analysed. § In conjunction specifically with point (c), Versuni notes that the tested material is a homogeneous densely woven fabric, for which the results of the machine used are considered reliable. This point was subsequently not rebutted by Dreame in its rejoinder. At the OH, Dreame asserted that the porous material used in the Mira Mop has a non- homogeneous chaotic structure and that, consequently, the test submitted by Versuni is unsuitable to measure its pore size. § In conjunction specifically with point (d), Versuni notes that the choice of the test liquid determines which pore sizes can be measured: a liquid with low surface tension, such as Porefil, can penetrate smaller pores than a liquid with high surface tension, such as water; for this reason, the two sets of measurements (with Porefil and with water) are intrinsically not comparable; the error associated with the mean pore size obtained from the Porefil measurements is at most 3%, which, according to Versuni, makes the test statistically significant. Dreame counters that the limitations in the choice of the wetting liquid for the measurements of pore size are below 1 μm and can be compensated by properly applying the equations given by the aforementioned standard ASTM F316-03. § In conjunction specifically with point (e), Versuni notes that the size of the sample (25 mm diameter) is due to the size of the sample holder; it was possible to take a 25 mm sample from the mopping pad of the attacked product, as shown in the pictures included in the Second Report; the tested area was free of glue. This point is not rebutted further by Dreame. 89. Moreover, in conjunction with the Counter Report provided by Dreame (Exhibit KAP6), Versuni notes that: § the test addressed therein is carried out according to standard GB/T 6165-2021, which doesn’t provide pore size, as it concerns the determination of filtration efficiency and 25 resistance of HEPA/ULPA filter media and filters when challenged with aerosol particles; as such, the test measures retention efficiency, which is not equitable to pore size; in fact, a material could have pores considerably larger than 0.5 μm while nevertheless capturing 57% of 0.3-0.5 μm particles; there is also no pore-size calculation anywhere in Dreame’s Counter Report; § section 6.1.1 of the standard used in the Counter Report establishes that no less than five filter-medium specimens are to be tested, and that the specimens are to have a minimum size of 200×200 mm (Exhibit VB-A41, p. 17), whereas the Counter Report indicates a sample quantity of one piece with a diameter of 50 mm. Subsequently, Dreame did not comment further on Versuni’s issues with the Counter Report. The objection by Versuni that the test used measures retention efficiency and not pore size, is therefore undisputed. 90. Dreame also requests that the Second Report be rejected as late filed. - The Court’s assessment 91. In view of the above, the Court will now consider whether there is a sufficient degree of certainty that the Mira Mop has this feature. It is recalled that Versuni has the burden of presentation and proof for facts establishing the infringement or imminent infringement of the patent, as well as for all other circumstances supporting its request. 92. The Counter Report filed by Dreame is discarded as irrelevant, because it does not provide information on the pore size. Consequently, the only evidence on file regarding the pore size of the Mira Mop’s porous material is the First Report submitted by Versuni. Dreame criticised this Report, including the method used. EP210 does not mention a specific method for measuring pore size. Therefore, it can be assumed that any method suitable for measuring pore size, known to the skilled person as accepted in the relevant common general knowledge, can be applied (see e.g. EPO case law on the issue, e.g. T1960/14). In this case the tests performed by a third- party laboratory, as documented in the Report, have been carried using a standard accepted in the field and considered relevant by the skilled person. Furthermore, the tests conditions are clearly explained when the clarifications provided in the Second Report are considered. The Second Report is admissible as it was filed in response to Dreame’s criticisms of the First Report and is not disproportionate or late filed. It is therefore taken into consideration. 93. Most of Dreame’s further criticism of the First Report was convincingly rebutted by Versuni, as summarised above. 94. Versuni has therefore fulfilled its burden of proof in the context of the application for provisional measures. In other words, it has satisfied the Court that it is more likely than not the size of the pores of the Mira Mop’s falls within the range of 1 μm and 50 μm. It then is up to Dreame to provide evidence which makes these findings questionable to the extent that the “more likely than not” assessment will turn in Dreame’s favour. 95. The Court considers that the evidence provided by Dreame is insufficient to call into doubt the results of the Report provided by Versuni, especially since Versuni has provided evidence about 26 the actual pore size, while Dreame has only provided evidence on retention efficiency, which is considered irrelevant for the assessment. To convince the Court otherwise, Dreame should have provided a full counter-analysis of the pore size of the porous material of its product. Dreame has not done so, although it had an opportunity to do so, and has not stated any specific reason for the absence of such proper analysis of their own, apart from asserting that the burden of proof is on Versuni. As explained, the Court finds that Versuni has fulfilled its burden of proof.9 96. Consequently, balancing probabilities, based on the evidence on file, the Court finds it more likely than not that Dreame’ Mira Mop fulfils all features of claim 1, including the pore size-range of feature 1.3. 97. As at least one Mira Mop was sold and delivered to a customer in Germany, an infringing act occurred within the Relevant Territory. It is further considered sufficiently substantiated that an imminent launch of the Mira Mop in the Relevant Territory can be expected, already illustrated by the fact that Dreame refused to confirm the contrary and requested a (royalty-free) license. Dreame with its existing distribution infrastructure in the Relevant Territory for similar products, can be considered to have set the stage for a launch of the Mira Mop in the Relevant Territory. III.G NECESSITY, RELIEF, BALANCE OF INTERESTS AND COSTS 98. Given the above findings on the likelihood of infringement and validity, the requested measures shall be granted insofar as they are necessary and proportionate (R. 206.2(c)–(d) RoP). In exercising its discretion, the Court shall balance the parties’ interests and consider the potential harm to each party if the injunction is granted or refused (Art. 62(2) UPCA and R. 211.3 RoP). 99. The provisionally established infringement warrants a general injunction. Versuni has convincingly explained that an injunction is necessary at this time. The issue of urgency has already been addressed and is assumed. Versuni is the market leader in its category of electrical mops, a fact which Dreame did not dispute. If Dreame’s Mira Mop enters the market, Versuni will suffer significant irreparable harm. Both products fall in the same category and target the same consumers. As these are durable consumer products with a long replacement cycle, they are in direct competition with each other. If Versuni is forced to wait for a decision on the merits, its market share is likely to be irreversibly eroded. This would result in irreparable and immediate harm that could not be adequately compensated for by damages awarded after a decision on the merits. Damages would include aftermarket sales of replacement parts and consumables, such as cleaning liquids. Contrary to Dreame's assertions, Versuni is not required to quantify or substantiate any specific risk of price erosion, price pressure, expected loss of sales or change in market share, even if this were possible, given that Dreame has not yet officially launched the Mira Mop in the relevant territory and therefore the sales price is unknown. 100. Dreame’s interest to deny the preliminary relief sought appears minimal. In fact, its own submission is that there is no threat of infringement in the Relevant Territory. While it was concluded above that this is not entirely correct (see 97 above), it can be deduced that an injunction would have minimal impact on its business. 9 Cf. CoA order of 3 March 2025, UPC_CoA_523/2024, Sumi Agro v Syngenta 27 101. The measures shall be granted with slight limitations as set out in the operative part of this order. Among other things, the requested penalties are limited and maximised. The parties have agreed on the costs, and this agreement is reflected in the operative part. 102. Dreame requested that a security for enforcement be imposed "in an amount to be determined by the Court, provided that the security shall amount to EUR 5,000,000". The wording of this request is unclear. It appears that the Court is being asked to set security for enforcement at an amount of EUR 5,000,000 or more. However, what the exact meaning is need not be determined, as the Court rejects the request to set security for enforcement. As the CoA held in Abbott v Sibio10, Dreame has not substantiated why serious difficulties would be expected in connection with the recovery of any possible damages from Versuni, a company based. in Europe. Furthermore, the request for security is inconsistent with Dreame’s assertion that they are not operating in the Relevant Territory. III.3 IN THE EP458-CASE (UPC-CFI-2595/2026) III.H THE PATENT, BACKGROUND AND TEACHING 103. EP458 was granted on 11 September 2024 for a ‘Cleaner head and wet cleaning apparatus comprising the same’. The application for EP210 was filed on 9 January 2023 (first published as WO2023/135092), claiming priority of 11 January 2022. EP458 proceeded to grant with substantially the same text and drawings as originally filed. No opposition was filed. 104. EP458 relates to a cleaning apparatus including a cleaning liquid supply and a dirt liquid retrieval function. 105. Wet cleaning apparatuses are known, which both apply a cleaning liquid onto a surface to be cleaned and remove the liquid (with dirt), e.g. by suction (paras. [0002]-[0003]). According to the patent disclosure, such prior art apparatuses have the following disadvantages: i. power consumption values for vacuum cleaners are relatively high, for example in the order of several hundred watts (para. [0003]); ii. providing both functionalities of liquid delivery and pick up can, in at least some designs, risk that the cleaning liquid is used inefficiently (para. [0004]); iii. poorly controlled delivery of the cleaning liquid may result in soaking the surface to be cleaned with such liquid; this soaking may not, in at least some circumstances, be easily addressed by the pick-up functionality of the apparatus, particularly when a relatively low power pick-up system is employed (para. [0005]); iv. in some designs, the pick-up functionality can also risk hindering the movement of a cleaner head of the apparatus over the wet surface to be cleaned (para. [0006]). 106. The teaching / objective problem underlying EP458 may be seen as that of providing a cleaner head for a wet cleaning apparatus that allows using the cleaning liquid more efficiency, requiring low power consumption (so as to keep into consideration also at least disadvantages (i) and (iii) mentioned above). 10 CoA 14 February 2025, Abbott v Sibio, UPC_CoA_382/2024, paragraph 170. 28 107. The solution to the above problem is provided (mainly) by feature 1.6 of claim 1, i.e. by said liquid pick-up region of the cleaner head being arranged relative to cleaning liquid outlet(s) so that the cleaning liquid being supplied onto the surface to be cleaned bypasses the liquid pick- up region. In this way, the cleaning liquid can reach the surface to be cleaned in a controlled manner and be used more efficiently. (Paras. [0018]-[0021]; [0212]-[0213]). In addition, the use of a porous material, preferably with a selected pore size, for said liquid pick-up region, with said material sealingly attached to dirt liquid inlets (feature 1.5/3.1), helps maintaining an under- pressure at the dirt inlets, thus reducing the overall power needed for liquid pick-up (para. [0013] et al). 108. A basic embodiment of a cleaner head (denoted by 100) is illustrated in Figs. 1 and 2 copied below, showing an underside (102) of the cleaner head – i.e. the side of the latter which, in use, faces the surface to be cleaned – and a cross-section of a cleaning liquid distribution strip (108) included in said cleaner head, respectively. 109. The cleaning liquid distribution strip (108) includes one or more cleaning liquid inlet(s) (112) in communication with a channel (110) which supplies the liquid to one or more cleaning liquid outlet(s) (104), arranged in a row parallel to a length (i.e. a direction of greater extension, according to the drawings) (106) of the cleaner head (paras. [0116]-[0118]). 110. Figs. 3 and 4 (copied below) refer to a variant embodiment wherein the cleaning liquid outlets (104) are arranged in two rows spaced apart along the width of the cleaner head and the porous material layer (114) is provided at a central location between said two rows (para. [0141]). 111. At a separate location along the width (118) of the cleaner head with respect to the cleaning liquid outlet(s) (104), there is provided a porous material (identified at 168 in the subsequent parts of the description and associated drawings), e.g. a porous fabric or a porous foam, including at least one layer (114). Below such layer (and therefore not visible in Fig. 1) there are 29 provided one or more dirt liquid inlets (two of which denoted by 142A and 142B in the subsequent parts of the specification, and visible, e.g., in the plan view of Fig. 5A and in the cross-section of Fig. 7B copied below). The dirt inlets (142A, 142B) may be defined by an opening of a tube or tubes (144A, 144B) (para. [0200]; Figs. 5A, 5B). The dirt inlets are fluidly connected or connectable to an under-pressure generator; therefore, upon application of a suction force by such generator, dirt liquid resting upon the surface to be cleaned is first transported through the pores of the porous material layer (114) and then passes from the latter into the dirt inlet(s). One or more further porous material layers (168) may be arranged between the layer (114) and the surface to be cleaned (see, e.g., the embodiments of Figs. 8 and 9, wherein such additional layers are denoted by 156 and 158). (Paras. [108]-[125]). 112. The porous material covering the dirt liquid inlets may contact the cleaning liquid applicator material (126, 128), allowing transferring some of the cleaning liquid to the porous material and into the dirt inlet(s). This configuration may assist in preventing an excess of cleaning liquid building up in the cleaning liquid applicator material (126, 128), thus minimising excessive wetting of the surface to be cleaned (as mentioned above in conjunction with drawback (iii)). Alternatively, or additionally, through said contact between the porous material and the cleaning liquid applicator material, the cleaning liquid in the latter may be used to efficiently rinse the porous material covering the dirt inlet(s). 113. With reference to the embodiments shown in Fig.s 6B and 7B (copied herein), a liquid pick- up region (PR) of the porous material layer (114) is delimited by a sealing attachment of said porous material layer around the dirt inlets (142A, 142B). 114. The sealing attachment prevents, or minimises, loss of the under-pressure via leakage between the dirt inlets and the porous material layer (114). Such sealing attachment can be obtained, e.g., by gluing or welding the porous material layer (114) around each of the dirt inlets (142A, 142B). Alternatively, an impermeable portion (146) e.g. a polymer film, sealed onto the porous material layer (114) and around the dirt inlets (142A, 142B) can be provided. (See, e.g., paras. [0014], [0206]-[0207]). 115. The dirt inlets can be defined by apertures in the impermeable portion and/or between the impermeable portion and the porous material layer (paras. [0044], [0094]). The dirt inlets may be exposed to a sealed cavity (150) defined between the porous material layer and said impermeable portion realizing the sealing attachment of the porous material layer around the dirt inlets. (Paras. [0205], [0206], [0209]; Fig. 6B). 30 116. In all the embodiments, the liquid pick-up region (PR) of the porous material layer (114) is arranged relative to the cleaning liquid outlets (104) so as to allow the cleaning liquid to bypass the liquid pick-up region to reach, or at least be directed towards, the surface to be cleaned (218). In particular, the liquid outlets are arranged in one or more dispensing parts which are spatially separated from the porous material layer (the arrows 240 in Fig. 24 copied below show the delivery direction of the cleaning liquid towards the surface to be cleaned). (Paras. [0368]- [0369]). 117. EP458 contains independent claim 1 and dependent claims 2 to 15 (only product claims, no method claims). Claim 1 can be divided into features as follows: 1.1 A cleaner head (100) for a wet cleaning apparatus, the cleaner head having: 1.2 at least one cleaning liquid outlet (104) through which cleaning liquid is deliverable; 1.3 at least one dirt inlet (142A, 142B) for receiving dirty liquid from a surface to be cleaned; and 1.4 a porous material (168) covering the at least one dirt inlet, the porous material comprising a porous material layer (114), 1.5 a liquid pick-up region (PR) of the porous material layer being delimited by sealing attachment of the porous material layer around the at least one dirt inlet, characterized in that 1.6 the liquid pick-up region (PR) is arranged relative to each of the at least one cleaning liquid outlet (104) such that the liquid pick-up region (PR) is bypassed by the cleaning liquid delivered towards the surface to be cleaned. III.I CLAIM CONSTRUCTION, SKILLED PERSON 118. Reference is made to part III.D above for the legal test for claim construction. Parties agree that the skilled person is an engineer involved in the design and development of household cleaning appliances, with a degree in mechanical engineering or product engineering and several years of practical experience. 31 119. The interpretation of features 1.5 is in dispute and relevant. Feature 1.5 – a liquid pick-up region (PR) of the porous material layer being delimited by sealing attachment of the porous material layer around the at least one dirt inlet 120. Feature 1.5 defines a (sub)portion of the porous material dedicated to dirt liquid pick-up and delimited by a sealing attachment of the porous material layer around the dirt inlets (para. [0044]). Such sealing attachment is consistently taught throughout EP458 as needed for maintaining an under-pressure at the dirt inlets (see paras. [0013], [0044], [0124]). In order to attain said technical effect – and as highlighted by Dreame – the skilled person understands the term ‘sealing attachment’ as requiring a structural sealing connection which prevents or substantially restricts fluid communication between the dirt inlet(s) and the environment external to the liquid pick-up region, as further explained below. 121. According to the patent disclosure, said sealing attachment can be implemented in various ways, such as by gluing or welding the porous material layer directly around each of the dirt inlets (para. [0014]; claim 13) or by using an intermediate element, namely an impermeable portion sealed onto a surface of the porous material layer exposed to the dirt inlet(s) and around the dirt inlet(s) (paras. [0014], [0206], [0209], [0211]; Figs. 6A, 6B, Fig. 7B; et al). According to EP458 the dirt inlets can be defined by apertures in the impermeable portion and/or between the impermeable portion and the porous material layer (paras. [0044], [0094]). 122. Based on said disclosure, the skilled person learns that the sealing attachment ‘around’ the dirt inlet(s) implies that such inlet(s) are encapsulated, i.e. enveloped, by the porous material layer or the impermeable structure. This can be affected by encapsulation of the dirt inlets separately or of all dirt inlets together. At odds with such construction appears the isolated statement at para. [0214] of the patent disclosure cited by Versuni, copied below (emphasis added): [0214] In other examples, the porous material can be attached, e.g. against the cleaner head 100 or a component of the cleaner head 100, around the dirt inlet(s) 142A, 142B at least partly by being sucked thereagainst by the flow provided by an under-pressure generator. Para. [0214] is introduced by previous para. [0210], also copied below (emphasis added). [0210] The impermeable portion 146 may, for instance, comprise, or consist of, a polymer film, such as a thermoplastic film. Various alternative sealing arrangements, some of which do not include such a polymer film, are described herein below. 123. Versuni considers the above passages as indicating that feature 1.5 is met also if no specific structural sealing means is used and the sealing occurs in operation only, when the under- pressure generator is activated and sucks the porous material against the dirt inlet. However, this construction is at odds with the technical effect consistently disclosed by EP458 in association with feature 1.5 – e.g. maintenance of an under-pressure in all operative conditions – and with the structural language of claim 1. Moreover, para. [0214] does not unambiguously disclose that a sealing attachment within the meaning of EP458 can be achieved by means of suction only. Therefore, Versuni’s construction of the literal meaning of feature 1.5 as possibly excluding a specific structural connection between the porous material and the dirt inlet(s) which envelopes the latter, is not followed. 32 III.J -INFRINGEMENT 124. Literal infringement of Inter alia independent claim 1 of EP458 is asserted with the Mira Mop. The allegedly infringing part of the Mira Mop is the (lower) cleaner head. The underside of the cleaner head has four cleaning liquid outlets arranged in a central row, substantially parallel to a length of the cleaner head, and a plurality of dirt liquid inlets arranged in two side rows parallel to said central row and obtained upon corresponding longitudinal protrusions (as exemplified in the two pictures below). 125. The underside of the cleaner head has detachable cover, the mop pad (see 5 above), covering said inlets and outlets. The mop pad is made of three layers, in particular a bottom blue microfiber cloth contacting the surface to be cleaned, a middle grey layer of porous material and a top blue layer of (impermeable) foam. A central part of the foam layer has apertures corresponding to the cleaning liquid outlets positioned there below. At lateral parts corresponding to the dirt liquid inlets, the foam layer has two elongated windows or apertures, which receive the respective protrusions of the cleaner head. The foam layer is glued to the porous material at the perimeter of said apertures (the mop pad is shown sectioned in the third picture below). 126. According to Versuni, while the grey porous material layer appears to have small holes, these do not align with the liquid inlet holes. The non-alignment of the holes in the grey porous material layer with the dirt inlets is evidenced, still according to Versuni, by the indentations in 33 the grey material layer matching the profile around the dirt inlets, as shown in the pictures below. Therefore, Versuni asserts, when the attachment is mounted onto the cleaner head, the porous material covers the liquid inlets. 127. Dreame contends the reproduction of features 1.5 and 1.6. 128. In the application, Versuni asserted that the entire space within each elongated aperture in the foam material – surrounded by the gluing of the porous material to this same foam structure – is the ‘pick-up region’ of feature 1.5 (see the picture at 125 above). The glue applied along the edges of the apertures is asserted to provide a seal around the dirt inlets (which are inside the apertures and against the grey porous material. After it became clear that the peripheral glue, shown in pink in the picture below, does not completely surround the apertures (glueless gaps are indicated with red circles), Versuni adapted its position. 34 129. It asserts that, independently of the adhesive bond, sealing attachment is also provided by the closed recess borders against which the porous layer is held when the pad is fitted. Thereto it relies on manufacturer’s documentation submitted with the reply as exhibit VB-B38, regarding a similar mopping device (Proscenic) originating from the same manufacturer as the Mira Mop. Relying on the picture below (obtained from exhibit VB-B38) it is explained that each liquid inlet is situated ‘in a racetrack-shaped recess that forms a sealed suction groove when fitted against’ the porous material layer. The ‘racetrack-shaped recesses’/protrusions within the apertures are shown in red. 130. At the OH, Versuni specified that the pick-up region includes the grooves defined by ‘racetrack recesses’ and that the sealing is obtained by the combined action of the peripheral gluing and the abutment of the borders of the racetrack recesses into the porous material. Versuni also reiterated that the fact that said gluing is partially discontinued around the perimeter of the elongated apertures is not relevant, as a ‘perfect seal’ is not required. 131. Dreame objects to the new evidence submitted with the reply both as late filed and as not sufficient to prove the correspondence of features between the Mira Mop and the Proscenic mop. Even if the evidence is considered, and if the regions immediately facing / adjacent to the protruding dirt inlets are regarded as the liquid pick-up regions, they are not delimited by a sealing attachment. The ‘racetrack recesses’ are rigid plastic structures that simply sit against the surface of the porous material layer and that do not provide any sealing. If the peripheral glue line is regarded as the relevant sealing attachment, it is not located around the dirt inlets and therefore does not delimit the liquid pick-up regions. 132. The Court is not convinced with a sufficient degree of certainty that the dirt inlets of the Mira Mop are enveloped by a ‘sealing attachment’ as required by feature 1.5 in any of the two constructs relied on by Versuni. It is not clear from Versuni’s submissions how such attachment, when construed correctly as set out above at claim construction, is implemented in the Mira Mop. It has not been shown that a seal is actually obtained at the elongated apertures of the attacked product with the peripheral gluing. Even if it is assumed that exhibit VB-B38 is admissible and that the ‘racetracks’ relied on are also present in the Mira Mop (which has not been demonstrated), Versuni has also not sufficiently demonstrated that the ‘racetracks’ 35 provide a ‘sealing attachment’ of the porous material around the dirt inlets as required by feature 1.5. It has thus not been demonstrated that this feature is implemented in the product. In view of this conclusion, it is not necessary to consider the (opposite) test results filed by the Parties concerning the implementation of feature 1.6. III.K – OUTCOME AND COSTS 133. As the Court if not satisfied with the required amount of certainty that EP458 is infringed, the request for provisional measures will be dismissed. It is not necessary in the context of these PI proceedings to decide on any other issues that are in dispute. Parties reached an agreement on the costs, which is reflected in the operative part. IV. ORDER Having heard the parties, the Court by way of provisional measures: A. IN CASE UPC_CFI_2594/2026 a) orders Dreame, within one week after service of this order, to refrain from infringing EP 3 169 210 in any way, including by making, offering, placing on the market and/or using and/or importing or storing for those purposes, the Mira Mop in the territories of the Netherlands, Germany and France; b) orders Dreame, within two weeks after service of this order, to provide to Versuni (p/a Brinkhof N.V.) a written account with information on: (1) the origin and distribution channels of the Mira Mop, or any other product that falls within the scope of protection of EP 3 169 210 within the territories of the Netherlands, Germany and France; (2) the identity, including full names and address details, of distributors, resellers, retailers and other legal entities involved in the sale, supply or distribution of the Mira Mop, or any other product that falls within the scope of protection of EP 3 169 210 within the territories of the Netherlands, Germany and France; c) orders Dreame to pay to the Court a penalty payment of EUR 10,000,-- for each day or part of a day that the granted injunction under a) is not complied with, and a penalty payment of EUR 1,000,-- for each day or part of a day that the order under b) is not complied with in full, such penalties to be determined by this Local Division upon request by Versuni, up to a maximum of EUR 500,000 in total; d) orders Dreame to bear the legal costs and expenses incurred by Versuni in the proceedings; e) confirms that the total amount of costs and expenses incurred by Versuni in this instance is EUR 225,000 as agreed by the parties; f) orders Dreame to pay to Versuni the sum of EUR 225.000 within two weeks after a written request thereto from Versuni; 36 g) declares the orders above directly enforceable; h) rejects the claims in all other respects; i) sets the date as referred to in R. 213.1 RoP at 30 calendar days after service of this order; j) sets the value of the case at EUR 5,000,000; B. IN CASE UPC_CFI_2595/2026 a) dismisses the Application; b) sets the value of the case at EUR 5,000,000; c) orders Versuni to bear the legal costs and expenses incurred by Dreame in the proceedings; d) confirms that the total amount of costs and expenses incurred by Dreame in this instance is EUR 225,000 as agreed by the parties; e) orders Versuni to pay to Dreame the sum of EUR 225.000 within two weeks from a detailed request thereto from Dreame; f) declares the order at e) directly enforceable. INFORMATION ABOUT APPEAL An appeal to this order may be brought in accordance with Art. 73 (2) (a) UPCA and R. 220.1 (c) and 224.1(b) RoP within 15 calendar days of the service of this order. (R. 211.6 RoP). Kokke Presiding judge and Judge-rapporteur Van Peursem Legally qualified judge Severinsson Legally qualified judge Papa Technically qualified judge On behalf to the registryHageman Larissa Donata 2026-10-26 15:23:03 +0100Kokke Margot Elsa 2026-10-22 11:53:57 +0200Van Peursem Gerard Robert Bernard 2026-10-22 10:21:58 +0200Elisabetta Papa Firmato digitalmente da Elisabetta Papa Data: 2026.10.21 15:33:59 +02'00'Sven Daniel Severinsson Digitally signed by Sven Daniel Severinsson Date: 2026.10.22 08:58:48 +02'00'

Key Holdings

• To obtain a preliminary injunction, the standard applied is whether infringement and patent validity are 'more likely than not'.
• As a general rule, a claimed solution must be considered obvious when the skilled person would take the next step prompted by a pointer in the realistic starting point or as a matter of routine to arrive at the claimed invention.
• A medical absorbing arrangement does not qualify as a 'surface cleaning device' under purposive claim interpretation, and the claim was not limited to liquid removal.
• Urgency is satisfied where the claimant acts diligently upon discovering the alleged infringement.

Tags

• Provisional measures
• Preliminary injunction
• Claim interpretation
• Inventive step
• Urgency