Written by Ben Ellison for Panbo, The Marine Electronics Hub
For me, this story began with a March 5th email titled “AIS patent wars – a tax on safety?” It referenced the SRT stock market announcement partially shown above and went on to say:
The whole point of using CSTDMA instead of SoTDMA in the original design of Class B was to avoid any problems with patents to ensure the successful uptake of the system by manufacturers. I see this as a tax on safety and a desperate attempt by SRT to force manufacturers to use their solutions rather than those from competitors and so create a pseudo-monopoly in the Class B world, which cannot be good for competition or the end user. To say I am furious would be putting it mildly…
The author of that email asked to remain anonymous, which is easy because I only know him as a nickname with an inscrutable email address who occasionally writes me with what seems like an insider’s knowledge of AIS technology and the regulatory process that made it happen. But anonymous recollections and opinions don’t have the weight of on-the-record testimony, and worse is “off the record” information which I can’t share at all. I ran into lots of both while researching this patent and that’s one reason why this entry will focus on documents that are publicly available or that I have seen with my own eyes. Readers will have to draw their own conclusions and the ultimate consequences seem murky anyway.
Patent 7512095 began as SIS
It’s easy for anyone to Google the full details of Patent 7512095. The invention is named a Multiple Access Communication System for Moveable Objects and AIS is not specifically referenced, nor are the SOTDMA and CSTDMA transmission techniques discussed in my recent entry about the “new” SOTDMA Class B standard. That may partially explain why several anonymous sources involved with AIS seem surprised that a CSTDMA related patent exists at all, though others are not.
It takes just a little more digging to find the 2003 Provisional Patent application that preceded the one granted in 2009. (Go to the official USPTO search site, fill in the CAPTCHA, and search for Application Number “60477125”.) Though still not called CSTDMA, this SIS patent concept is clearly about the Carrier Sense TDMA technique that replaced SOTDMA as the first Class B standard. In fact, the necessary “shorter packet” explains why a new two-part Class B static data message had to be created, which went on to cause a number of issues (as also discussed last week). Note, too, the first paragraph reference to the “patent issues which impact cost” {of the then current SOTMA AIS method}. While the differences between the 2003 patent application and the 2009 patent are great, be careful about drawing conclusions; I’m told that it’s common for provisional patent language to be broadened considerably in the final draft.
The IEC connection
I’ve never met the original patent applicant, Mark Johnson, but know him to be the proprietor of Shine Micro and purportedly the engineer behind some of the best AIS receivers on the planet. He’s also a longtime member of IEC TC 80/WG 15 — the AIS working group of the IEC’s subcommittee on maritime navigation and radiocommunication equipment and systems — as is Andreas Lesch, Johnson’s co-inventor for patent 7512095 (who apparently sold his share to SRT last March). Therein lies something of a mystery given the often heard notion that TC 80/WG 15 developed alternative CSTDMA to avoid patent restrictions.
I have seen some credible documentation of the Class B/CS standards-making process, including what seems to be a long email chain amongst WG 15 members. It seems clear, and it’s acknowledged by many, that Mark Johnson took a lead in developing the CSTDMA standard, writing on June 7, 2003, “I will be proposing a non-SOTDMA lower cost Class B variant that is low in cost, very ‘polite’, and compatible with existing Class A AIS. I am quite confident that it avoids the IPR issues.” That was one day before he filed the SIS provisional intellectual property claim referenced above, but again be careful with conclusions. One anonymous explanation of this apparent inconsistency is that the patent was filed only as a potential defense if some other entity claimed CSTDMA rights.
The IEC has a clearly stated policy that anyone participating in the work should reveal any known patent or pending patent application, either their own or of other organizations, and there’s a form for the purpose. If you go to the IEC’s patent declarations database and search on IEC Class B AIS Standard “62287” you will only find one declaration, partially shown above. It was filed by Anders Håkan Lans in 2007, stating that he would license manufacturers of SOTDMA Class B on a non-discriminatory basis with reasonable terms and conditions, though it seems that the IEC did not complete the B/SO standard until after the Håkan Lans patent claim was cancelled after reexamination. RAND or FRAND licensing terms are acceptable to the IEC, however, though the “free of charge” option #1 above is probably preferable, and patents under option #3 are not supposed to be in IEC standards.
All of this may help to understand SRT’s stance on the patent and also the recollections of the one TC 80/WG 15 participant willing to go on-the-record so far. That’s Joe Hersey, Jr., who is now retired from a 27-year USCG career deeply involved with telecommunications and standards-making, but who remains secretary of the US National Committee’s Technical Advisory Group to IEC TC80. Here’s his answer to my query about the Johnson/IEC patent situation back in 2003:
Yes, I believe that the IEC and its Working Group 15 (responsible at that time for Class A and B) were very well aware early in the CSTDMA development process that Mark had filed a provisional patent application, although I cannot speak to exactly when they became aware. Mark had made it very clear from the beginning of development of the standard that he had done so, but that its use was being offered free of charge on a non-discriminatory basis. The Working Group was very much aware of the patent difficulties associated with the Class A device and wanted to ensure the Class B avoided such difficulties. Since IEC could not complete a Class B CSTDMA standard until the technique were recognized in ITU-R Rec M.1371, the IALA AIS technical working group, had to prepare a CSDTMA amendment to that Recommendation too. The IALA AIS TWG members were also aware of the pending patent application, and that its use was being offered free of charge on a non-discriminatory basis.
And here’s Hersey’s response to my question about the possible effects of SRT’s licensing plan:
Because this patent has been long known and understood by the AIS standards committee members to be offered free of charge on a non-discriminatory basis, I question how SRT’s plan to suddenly offer licenses on a FRAND basis could be beneficial to the AIS industry and consumers. IEC, ITU-R, ITU-T and ISO have long held a common patent policy that requires participants in standards committees who hold a patent which may affect a standard under development to file a statement indicating which of three conditions licensing would be offered: free of charge and non-discriminatory, non-discriminatory and reasonable terms & conditions, or neither. If neither, the standard could not include provisions depending upon that patent. Hakan Lans filed noting the second option under the Class B SOTDMA standard IEC 62287-2. No filing was ever made regarding the Class B CSTDMA standard, despite Mark Johnson and Andreas Lesch (the later co-owner of the patent) being intimately involved in the standard’s development.
SRT on the record
I first looked into the IEC connection back in March and it happened that CEO Simon Tucker gave me SRT’s official take on it just after WG 15 had used its facility for a meeting where the patent was briefly discussed. You’ll see that Tucker had much more to say about why SRT acquired the patent and why their management of it may be a benefit to the industry and consumers:
This Patent and its founding prior-art IPR was known to the relevant IEC committees for many years, including before and during the development of the CSTDMA based Class B standard. However no formal license statement and or license procedure for this patent had been established by the owners. SRT was the first company to develop a low cost Class B and today offers a full range of AIS products from AIS Aids to Navigation and Class A to Class B and Coast station technology and product solutions, each of which offers best in class functionality and reliability. Our investment in leading edge technologies and robust and reliable derivative products has seen AIS products be substantially reduced in size and cost, simultaneously with their functionality and performance substantially increased. SRT has a long term commitment to AIS and is thus, of all companies, interested in its continued development within the correct standards process as defined by the IEC committee. As a responsible company SRT took the opportunity to acquire rights to the Patent to secure its use for the AIS standard going forward. Last week {March 20}, SRT hosted the IEC standards committee meeting at which this topic was discussed. The IEC AIS standards committee has a clear process for the licensing of IPR / Patents which are required for technical standards created by the IEC, and SRT confirmed that, as a responsible international company, that it would be following this process and therefore offering licenses for the Patent on a FRAND basis. This action by SRT ensures that a critical part of the AIS standard is secure for the benefit of all AIS stakeholders for the future and normalizes its availability. This is great news for the long term stability and continued evolution of AIS technology and products.
Regardless of the patent’s history and its possible safety in SRT’s hands, the underlying concern I first heard — “A tax on safety?” — remains, at least for some. Here is Simon Tucker’s recent response on that subject:
I see no valid reason why the licensing of a long known and established patent which ensures a technology standard works well should affect end user pricing; proof of this is that licenses are already agreed and there has been no price increase. Essential and other IPR licensing into products is a normal procedure in most products and industries these days – the average mobile phone or drug can have up to 20 of these. It’s down to the manufacturer to keep pace through continuous investment in R&D to generate more value in their product than the cost of the third party IPR they wish and or have to use. In our view this is a natural competitive process which in the end ensures that real R&D continues, and that companies seeking to provide the market with products never become lazy. Of course those companies who have chosen not to invest in new core technologies and products for years on end, take their profits and run, will not want or like change. But in my view the evolving requirements of the market are to be respected at all times and it expects manufacturers to continuously invest in order to win their business.
As for actual license details, an anonymous and displeased manufacturer showed me an email that appeared to originate from SRT and proposed “a quick and amicable agreement without publicity or any disruption to your business” with the following terms: “….there are well set precedents/norms for essential IPR such as this which is 5% of the net of tax retail price. We estimate that the average Class B retail price is in the region of $700, thus indicating a per unit royalty of $35 per unit. Our legal right to the per unit royalty extends to all sales over the last 6 years prior to our first expression of rights, and then going forward in the future. Thus, there would be a lump sum in regard to the arrears and then quarterly payments on sales until March 2019 when the patent expires.”
On the other hand, another anonymous AIS manufacturer responded to the initial SRT patent news with: “I don’t think it is a major issue for us. I do think they are trying to protect themselves from an onslaught of $200 Class Bs from China. In that sense it should help us as well.” There are at least six independent Class B AIS manufacturers: Garmin, Furuno, Icom, AMEC, Weatherdock, and Vesper Marine.
Vesper Marine challenges the patent
Finally, there’s the question of Patent 7512095’s validity. In late April, a Seattle law firm representing Vesper Marine asked the U.S. Patent Office to reexamine all 14 of its claims against what’s known as prior art. The 124 page petition and many related documents can be viewed or downloaded at the USPTO search site (search for Application Number “90/013498”) and may in themselves be an indicator of the industry costs in such circumstances.
Here are Joe Hersey’s thoughts on the patent’s validity:
I am uncertain whether the patent is a valid claim to an original invention for a number of reasons. I believe some question to this effect existed at the time, but because it was understood that the patent was being offered free of charge, it was not considered to be a concern. For example, Mark did not invent the CSTDMA technique alone but did so in cooperation with the USCG R&D Center, David Pietraszewski in particular. The Coast Guard has, in my opinion, an equally legitimate claim to the invention of the CSTDMA technique. Secondly, the “listen before transmit” technique central to CSTDMA slot access is not uncommon in telecommunications technology.
In early June, the Patent Office announced its determination that “A substantial new question of patentability affecting claims 1-14 of the ‘095 Patent is raised by the Third Party’s request…” and that the reexamination will proceed (as shown below).
So that’s the situation as well as I can report it at this time. Of course, I welcome corrections and commentary. The reexamination process as I understand it is that SRT has an opportunity to respond to the reexamination request, the Vesper Marine representatives may respond to their response, and then the USPTO will make a judgement on the patent’s validity. That’s why I wrote that the consequences are murky. The patent may affect AIS developers and users in some ways until 2019 or not. But what’s happened so far has happened, confusing and complex though it may be, and there may be consequences for the companies and individuals involved. The floor is open.
