with Istvan Sebestyen from the European Computer Manufacturers Association (ECMA)
The following interview with Dr. Ivstan Sebestyen held in April 13th 2010 was conducted in the context of a fact finding study about the “Interplay of IPR and Standards”. Ivstan Sebestyen has been involved in the worldwide multimedia standardization work for over 20 years including telecommunication standardization experience in CCITT, ITU-T, ISO/IEC, ETSI and DIN and ITU-T Rapporteur on still picture coding (JPEG, JBIG) between the years 1987-2000. He has been involved in IMTC work since its very beginning in 1994. Since 1999 he served the IMTC as Director, and from 2001-2006, was elected as the President of the IMTC. Istvan Sebestyen has written close to one hundred papers, two books and holds close to twelve patents. He is now the President of ECMA International, an industry association dedicated to the standardization of Information and Communication Technology (ICT) and Consumer Electronics (CE).
1. How often (please quantify) are IPRs, especially patents, included in standards produced by your standardization body?
We (ECMA) have about 400 standards and I would estimate that in between 5-10% of these standards include IPR. This rather low percentage is connected to the technical categories. If you take our standards in DVD or CD you find lots of IPR especially patents. Theses patents are only owned by a little number of patentees though. But if you take for example programming languages (such as java, fortran, cobol) you will not find any IPR. It does not only depend on the technology but further more on the industry. Products for the internet market usually have no patents and if so, they are usually royalty free.
2. How often is this the result of active contribution of technology by members and how often are patents from third parties claimed essential to standards (either by declaration to the SDO or post-standardization infringement claims to implementers)?
We do have third party organizations that hold patents on our standards. This is a very difficult field, because we do not always know who they are and which patents they own.
3. In which technologies does this take place? Are there also specific technological areas that have standards with little to no patent claims?
I would say in most cases this happens in the field of multimedia and codex standards, where you find very complex IPR situations. In these technological areas many companies (often 40-50) hold patents and thus try to introduce the patented technology in a standard. But if you look at the ECMA standards these cases are very few, which is due to our standards that are usually not in these technology fields.
4. What companies (small/large, (non-)standard implementers) and other organizations (research institutes) contribute essential IPRs to standards?
Well I would say this is quite heterogeneous. I guess this depends more on the standard, since we have organizations that act differently between the standards projects. Microsoft for example holds essential IPRs in some standards but also works in groups without IPR protected technologies. And even the licenses for the IPR protected standards differ within one company and they participate in different project that are royalty free or RAND.
There are some companies like for example Philips, which receive high amounts of royalties from their IPRs in standards. Same with a company like for example QUALCOMM which makes profit from patents in the standards but not much from the product itself. But these players can usually only be found in certain technology fields such as speech codex (no standards in ECMA). Research institutions such as Fraunhofer or usually offsets of Universities can also be found. But as I said not much in the ECMA consortium, but when I think back to my time at ITU there was for example the University of Sherbrooke that holds patents on speech codex standards. I guess the patent holding players may appear very heterogenous.
1. Could you specify the impacts from integrating IPRs in standards? Are related standardization processes hampered?
This again is dependent on the project and differs between the standards. We do have projects where companies introduce innovative technologies, which are IPR protected and these inputs benefit the standard. But there are also cases where companies give inputs (IPR protected) which do not deliver useful technology for the standard. These companies would only agree to a certain standard, if they are allowed to integrate their technology. You can actually call this blackmailing if you like. Other members have to decide whether or not to accept these conditions (blackmailing tax).
When I worked for ITU we had the case of specific jpg codex standardization. IBM was the company who actually invented the product. Other companies like AT&T and some other Japanese companies joined the project to only introduce non relevant IPR protected technologies. Again they were only willing to agree to the standard, when their technology was also included. This makes the standardization process more complex and time consuming and sometimes even introduces errors on products.
2. What are the implications on the implementation of standards, and is the situation different according to types IPR licensing policies?
The standard setting process and the licensing process are strictly divided at ECMA and I would even say at most standardization bodies. The companies state, which IPR is included in the standardized technology. This is on a voluntary basis. We are also not able to verify whether the company’s IPR statement is true or not. All we can do is to publish these patent disclosures in our system so that everyone is able to access the information. ECMA only checks the formal part of the included information and we only publish what is communicated to us. Implementing organizations thus have to approach the patentees theirselves, if they want to use the standard technology. IPR that is not communicated to us (mostly from third parties) is not published but theses patentees will also demand royalties if the technology is used. There is nothing we can do about it and there are also some litigation cases happening.
3. Is there competition between standardization bodies based on different IPR policies?
ECMA has been marked conform right from the beginning (to ITU, ISO,IEC, ETSI, IEEE). ECMA was actually one of the first who had a patent policy. In December 2009 we actually rewrote the IP policy and now it is even closer to the other above mentioned standard bodies. But we still have ECMA specific clauses. Our IP policies are far from perfect. Especially the strict separation from standard activities and licensing is not optimal. We do not receive a feedback from implementers and so we are not able to build up an IP policy that considers practical experiences. This is a common problem at all standard bodies. I think in the future we will have to address these problems in more detail.
1. Are IPRs blocking the initiation of standardization processes? How does the screening take place, both of those stakeholders participating in the standardization process, as well as of those not participating but owning relevant patents?
This is again different in every project and I can identify three different procedures.
– There are some projects, where we do not have a screening at all. The standard process just starts and in the end all participants are requested to state, whether or not they have IPR that covers the technology. So right before the standard is published all the companies should make an IPR statement but not before.
– I know from other projects that there is quite a strict screening before and during the standard process, where every input or request for an input has to have an IPR statement. Again this is mostly on a voluntary basis. I know that the video group of ITU has such arrangements. Every contribution of a company needs to have an IPR statement at hand. We don’t have these arrangements at ECAM since yet, but this is possible for future projects.
– The third case is for example a programming language. At ECMA we have the ECMA script standard, which is a standard for java script. This programming language will be used in the internet and all participants know from the beginning that there should be no patents or royalties that cover the standard. And if there are IPRs on some technologies all members agree from the beginning to a royalty free policy.
We also had ex ante licensing (prices) in some projects and this is a practice which we are following since 20 years, but only for some projects. Examples are the jpg project or the standardization of the arithmetic codex at ITU. That was the example with IBM and AT&T I mentioned earlier. There we had a statement that the licenses should not exceed 5.000$. This was 20 years ago. Interestingly was, that this standard was later on given to Lucent and they did not comply to the 5.000$ statement and took 50.000$. These cases are quite rare though.
2. How do you deal with IPRs which are applied for or granted during the standardization process? What role does the fast track procedure play?
The special situation with fast track patents on standards is that everything has to be handed in twice (IPR statement). So the companies have to state patents at ECMA and ISO. But the problem is that there is no coordination or harmonization. So the patent statements for one standard may differ between the standard bodies. But this problem also occurs at JTC1, where IEC and ISO are not able to harmonize their patent statements. ISO and JTC1 are also having projects with ITU and all these organizations have different patent statements and each time the process has to be done again. This problem is a big future topic to be solved.
For example we had a standard developed with ISO, where we had a patent statement each time (for the ECMA and the ISO version) but partly about different patents from different patentees. ECMA and ISO are not able check this. We only receive the statements and present it, but we do not communicate/coordinate with ISO about these statements.
1. Do you have a specific IPR policy? What are its main features and characteristics? Do you have guides for the implementation of the policy? What is the position of your organization with respect to transparency, conflict resolution and licensing, e.g. FRAND?
This is totally dependent on the produced standards and differs between the standards groups within ECMA. There are different business cases for each project where we identify the technological field, the sector and the participating organization. If we work with W3C (ECMA script) we already know that the IPR will be royalty free since they have this policy. We do have standards that include IPR, which are royalty free, others demand for royalties which have to be RAND licensed. We have IPR policies for patents (usually FRAND) and copyrights, trademarks are not described yet.
2. Does your IPR policy address such software specific issues (e.g. copyright policy related to software) or do you plan to do this in the next 5 years?
Right now we work on the software policy topic. Especially for running software we need certain arrangements to identify what the functions of the products are and that ECMA is not responsible. So we are currently drafting an experimental copyright policy for one of our standards groups. After that we will probably create a general policy or will write policies for each project. This will be finished this June.
3. How do you understand the interaction between such implementation under open source licenses and the essential patents related to such standards? Do you have experienced issues in this field? Please describe?
We have one project, where we want to offer an open source license. This is due to the nature of the product though. It will be the next generation of ECMA script. We are still discussing which open source license to take. I think the BSD license (Berkley Software Distribution) will be the most suitable. The problem with BSD is that this license has a patent component. We would prefer that this component will be the ECMA IPR policy and right now we want to establish such a solution. So yes we are actually searching for open source solutions. But I think in the future we will offer different licenses depending on the case. We also need licenses not only for the end product, but also for the standard developing process which could be an open source license.
4. What other issues concerning the interplay off standards and IPR have to be addressed from a standardization body’s view? What solutions do you propose, and by whom? Do you propose further institutional solutions or changes of the current policies and their implementation?
In the future, patents will be the most difficult problem to be solved. In the field of multimedia, DSL, ADSL or RFID we find a concentration of a high amount of patents. But this is just my practical experience. I think it will be interesting to measure the concentration of patents in certain technological classes to identify the problems more easily.
Copyrights and software is a field, which has to be approached in the next years. ITU has a copyright policy but it is rather weak. Right now ETSI, ECMA and IEEE are planning to create such policies but ISO and IEC are not planning to adress these issues yet. I still think all this will be developed during the next 5 years. Also trademarks should be approached with some guidelines as ITU already did it.
Writing a good IP policy is a difficult job and I think there is not always a clear solution that incorporates all issues.