Over the last year, different groups have been beavering away to decide on a wireless standard for smart meters. It’s been interesting to observe the ways that different countries have approached this. There’s been the pragmatic approach of going with what’s available today, but with the understanding that it might need to be changed, so that everything currently being installed is at risk of needing replacement. That’s the UK approach of DECC. Then there’s the academic approach which is favoured by SGIP in the US, which entails producing a giant matrix of the vital (and not so vital) statistics of every possible wireless standard. At which point there will presumably be a flash of smoke, a glamorous assistant and a magician will be brought on stage to perform the conjuring trick of comparing apples, lobsters and elephants and deciding which is most appropriate of them for the smart energy feast. Or we have the slightly nepotistic ETSI approach over in Europe, which seems to be one of giving EU funding to all of their consultant or professor friends, who in return for this largesse promise to write their own, brand new wireless specification in time for the party.
Whilst some of these approaches consider cost in terms of the price of silicon, or even the opportunity cost in terms of time to market, one significant cost has been missing from their calculations – the cost of choosing a standard that opens up Intellectual Property disputes. That’s a real risk. The only place I’ve seen it publicly stated is in a briefing document from the Bluetooth SIG, which points out that from the IP viewpoint, wireless standards are far from equal. It’s a very valid concern. We’re already seeing the patent trolls coming out and attacking ZigBee and Wi-Fi. As volumes start to increase, so will their determination to make a fast buck. As soon as that happens, deployment could grind to a halt.
Intellectual Property Rights in standards is something that most people ignore. There’s an assumption that if a standard exists, then as long as you acquire the right to use it, there are no IP issues. Unfortunately that’s not true. A lot of companies have been taken to court for that misplaced belief, and ended up paying millions of dollars, or in extreme cases, going to the wall. So it’s important to understand the IPR landscape with respect to standards. I’ll start with the same statement that’s in my book:
The existence of a standard does not give a company a free right to exploit it, nor automatic access to the IP within it. To use it you must agree to the terms of the body which owns the standard.
These terms vary, but if we look at ZigBee and Bluetooth, each requires a company that uses the standard in their product to sign a member agreement which contains a clause of IP ownership. You can read the Bluetooth one here and the ZigBee on in Annex 2 of their agreement. Both introduce the concept of Necessary Claims. What that means is that once you sign up, which you must do if you want to use their technology, you effectively give up your right to sue for infringement of any IPR you might own which is included in or used by the other companies who are implementing the standard. It doesn’t affect you rights if it’s used by companies outside the standard, or by companies who use the standard, but haven’t signed that agreement – you can still sue them. In Bluetooth you can’t charge a fee for anyone using your IPR as part of the standard – that’s an RANDZ license (Reasonable and Non Discriminatory – Zero cost). In ZigBee, in theory you can charge a license to other ZigBee companies, although I don’t think anyone has – that’s a RAND licence (Reasonable and Non Discriminatory, with the emphasis on reasonable).
This is a two way pact. Companies agree to give up the right to sue each other, but in turn sign away some of the value of their IP. The latter can be quite a difficult step for some companies to take. The use of IP as a competitive tool varies quite widely between different industries. Some see it as having a defensive value, some as an aggressive means of getting income, other as a bargaining tool. When you sign this sort of agreement, you’re also signing yourself up for potential future use of your IP that you might consider valuable, because no-one knows what these standards might cover in five or ten years’ time. That’s something that might make companies with a more conservative approach to IP reluctant to sign.
The benefit of this model is that the more companies you can get to sign the IP license, the more IP goes into the standards pool, so the more protection you’re likely to have from being sued. Bluetooth is particularly good in that respect, as they have over 13,000 members who have signed up. ZigBee has around 300. Before a standard is released, most standards groups do an IP review to see if there are any problems from patents held outside the group and will try to address them. But none of the standards bodies give any IP guarantees. If there’s an underlying patent that someone owns, they can come out of the woodwork and sue you. No standard is immune – that’s already happened to Bluetooth, Wi-Fi and ZigBee. The question that companies and organisations need to ask is how to compare the risks?
To start, you need to look at just how much a standard covers. Some standards, like Bluetooth, have written everything from the radio specification up to the application profiles. So all elements of that specification are covered by the member IP agreement. Others, like ZigBee, use external standards for the radio and baseband (in their case the IEEE 802.15.4 specification), so the ZigBee agreement provides no IP cover for any of this.
Even when a standard has written the whole specification itself there are still subtleties. Z-Wave is a standard that was written to cover the whole stack from radio to application. But it was written and the IP is owned by one company. It means that it does not have the safety of 13,000 contributing members for its IP pool which Bluetooth has. On the other hand, it’s a simpler specification, so there’s probably less to infringe. You don’t really know the answer to this until there are enough products in the market to make it worth an IP owner taking out an injunction. Bluetooth got past that point a long time ago, as there are billions of Bluetooth devices in the market. Standards like Z-Wave haven’t shipped enough to make it worth a patent holder suing them and ZigBee is only just at that threshold.
Which brings us to patent trolls. Many people ask “What is a patent troll?” Within the PC and telecoms industry it’s long been rumoured that some large companies employ more lawyers than engineers and make more from their patent licenses than they do from selling products. Whether you think that’s a good business model is up to you, but at least they’re generally trading on their own innovation. Patent troll is the term applied to companies that specialise in buying up unused patents – often from failed start-up companies, sitting on them until they see a market opportunity and then suing companies who infringe them. Because they don’t generate their own IP, many in the industry see them as parasites – hence their derogatory name. (For a nice example of some of the more inventive IP business models, I’d recommend Cory Doctorow’s novel Makers.)
Here is where the uncertainty about the risks come in. The business model for someone owning some underlying IP is not to go after the first product to appear. If you do that you may frighten the market off, which would limit your licence revenue. Instead, you wait until the technology is well established, probably shipping a hundred million units a year and then send out the injunctions. Typically these will be served against smaller players who can’t afford the legal battle, or the cash-flow hit of having their production stopped. If they take out a licence, this will be used as legal proof that the industry considers the patent is valid, at which point injunctions will get slapped on the bigger players.
That approach is taken both by trolls and research institutes that own relevant IP. It means that the risks and costs of choosing a particular standard may not be felt until four or five years after that decision was made. As a general rule, it means that standards that have been around longer, have more members and have also shipped more units are safer. Once a standard has passed the hundred million shipments, most of the trolls will have come out of the woodwork, so you can be pretty clear about the future risks. If it hasn’t, then you really need to do your homework, and probably put aside five to ten dollars licence contingency for every product you make. Which is a cost that is generally left out of the PR pitches made by standards bodies.
The corollary is that it also means that more recent standards are more risky. People owning IP that they think they can make money from keep quiet about it. And as each year passes, more patents are being granted in this area, increasing the threat for new standards currently being written. So unless you have the wide membership of a standard like Bluetooth, any new standard, protocol or even profile offers the risk of infringing an ever growing body of IPR.
Any patent owner in this area will have realised by now that they have the potential to put the smart metering industry over a barrel. The cost of physical deployment is so high, that any infringement will have to end up with a settlement – it’s too costly to replace the wireless portion of millions of meters. The question those specifying the standards need to ask themselves is whether they’ve fully understand that risk when they make their choices.