Context: This is the fourth article in a series on the issues surrounding the proposed EU regulation on standard-essential patents (SEPs). The objective is to explain basic concepts as well as licensing and litigation realities as a contribution to the debate.
Previous parts: Part 1: Hold-up vs. hold-out; Part 2: EU-wide law because of German rulings?; Part 3: Small and medium-sized enterprises’ problems not SEP-specific.
In yesterday’s news on Nokia v. Verifone (March 14, 2024 ip fray article followed up by a discussion of the five patents-in-suit and related strategy), ip fray already noted that this case has no implications with respect to the EU SEP Regulation. Verifone is a multi-billion-dollar U.S. company, not a European startup.
This continuation of the series on the EU SEP Regulation was planned anyway, and the term IoT came up in previous parts of the series, but against the backdrop of Nokia v. Verifone and the way some have already tried to link it to the EU policy debate it’s even more pertinent.
Just like “SME” (part 3 of this series), the term is overused because of its appeal to politicians’ protective instincts. Internet of Things is a term without clear boundaries. Even with a reasonable delineation, it still contains disparate product categories that are lumped together this way, but actually have distinct characteristics when it comes to SEP licensing.
IoT potentially means any internet-connected object that is not a traditional computer or mobile computer (smartphone, tablet)
The history of the term is that initially the internet was a network of computers that can send and receive data over it, with some being effectively servers and others being effectively or mostly clients. The two primary applications in the early days were email and, from the mid 1990s thanks to Netscape and Internet Explorer, web browsing. Just two days ago, the inventor of the World Wide Web, Tim Berners-Lee, published a Medium post on the occasion of the 35th birthday of the web and rightly expressing his concerns over the web being exploited by corporations locking in users (without mentioning any, it’s a safe assumption that Apple is among them and likely number one on the list).
The web was invented at a time when mobile telephones, even based on the digital GSM standard existed, but it took a while before phones connected to the internet. The first BlackBerry email pager wasn’t launched until late 1999.
By the mid 2000s, more and more phones (not yet the cheaper models) provided some internet functionality (web and mail), but it wasn’t really until the launch of the iPhone in 2007 (followed shortly by the first Android phones) that the mobile web became a mass market phenomenon. Screen size was the primary bottleneck, which Steve Jobs identified as a bigger problem than the inconvenience of a touchscreen versus an actual keyboard (BlackBerry devices had physical keyboards).
No one argued at the time that the smartphone was the beginning of IoT. Instead, Apple and others viewed the smartphone as a mobile computer, given that you can do most of the things on a smartphone that you can do on a desktop computer, though for some applications such as video editing you really want to have a huge screen.
By now, every internet-connected object that is not a traditional computer or a mobile computer (smartphone or tablet) is described as an IoT product: not by everyone, but for every connected product you’ll find people who will call it an IoT product.
It’s a highly heterogeneous concept.
From cheap and small to large and expensive: asset trackers and airplanes
At the low end of the range, you find asset trackers that are just stickers. All they do is transmit location information. Unlike a printed bar code that will only serve its tracking purpose if someone reads it, such trackers are capable of sending location information themselves.
The most extreme examples at the other end would be space stations, intelligent (highly automated) buildings, or (if nothing else) airplanes.
In between you have a huge range of devices, but also all sorts of vehicles (passenger cars, trucks, agricultural machines or military equipment).
In that sense, an IoT-related argument was made in the dispute between Microsoft and Motorola over WiFi and video codec SEP royalties in the early 2010s. Motorola, which didn’t really want to sell a license but instead sought to force Microsoft into a zero-zero cross-license agreement with itself or (after being acquired by Google) with the entire Android ecosystem, made a prohibitive demand: they wanted 2.25% of the price of an end device, and they specifically explained in a letter that even though Microsoft merely sold Windows to computer makers, they’d want 2.25% of the price of a laptop. To explain how unworkable and unreasonable that proposition is, Microsoft gave examples of how much it would then cost if they collected their WiFi royalties on a $100K car (that per-unit royalty would pay for plenty of WiFi routers) or on a large passenger airplane.
The role of connectivity ranges from nice-to-have to very useful to essential
Not only is there a huge price range and are the relevant products sold in a diversity of markets, but there is also the issue that connectivity (increasingly cellular, but sometimes still just WiFi) makes a contribution that varies greatly from product category to product category:
- A product may not actually be better-suited to serve its purpose just because it’s connected, but connectivity may still come in handy for maintenance purposes, often including the possibility of over-the-air software updates. Think of some manufacturing equipment that is installed in a truck so it can be moved around: its job is to manufacture something, and it doesn’t crank out products better or faster owing to connectivity, but if there’s an issue, it’s nice to be able to contact tech support and let them perform some remote diagnosis or maintenance.
- A product may overwhelmingly serve another purpose, but connectivity is appreciated. Airplanes are an example. Primarily their job is to take people from A to B, but especially on a long flight, some passengers, especially but not only business travelers, will want to connect to the internet.
- A product may exist without connectivity, but it becomes better, possibly even a whole lot more useful. Smart meters can perform their basic metering function anyway, but the practical difference between a connected one and one that requires periodic reporting is huge.
- A product may essentially be enabled by connectivity (such as asset trackers that must work in places where there’s no bar code reader around; think of cattle running around on a huge piece of land). It wouldn’t exist without it, at least not in any comparable form.
Verifone’s payment terminals are used for a purpose that requires at least some sort of connectivity. Otherwise one could just use one of those credit card imprinters that were in use until the early 2000s. That connectivity doesn’t necessarily have to be wireless. It can also be connected by cable, such as a cashier desk. But you need at least WiFi if you want the waiter to bring it to your table. And if a merchant has to be location-independent in an area of a certain size, cellular connectivity is the only practical option.
So even within the product category of payment terminals you can see that there is a difference between those that can do with WiFi and those that need cellular connectivity, with a gray area in between where WiFi would work but cellular is simply preferred.
Cars are also an interesting example. Connectivity enhances the value of vehicles in various ways, and autonomous driving will be the ultimate goal (at which point people will also want connectivity to make use of all of that freed-up time). Connectivity is also used for customer service purposes. And let’s not forget that connectivity is used to generate incremental revenues (annual subscription charges or feature activations).
Royalty rates, royalty bases and licensing levels can therefore vary greatly
When you have such a heterogeneous concept as IoT, it’s obviously impossible to determine a one-size-fits-all royalty rate. What works for an asset tracker won’t be fair for a car or airplane (not even for a payment terminal, actually).
One company that proponents of the EU SEP Regulation sometimes portray as part of the problem is actually part of (and even a driver of) the solution in IoT SEP licensing: Huawei. On its website, the company lays out a highly differentiated royalty structure for IoT purposes. That is an example one should encourage others to follow.
There are different schools of thought as to how different IoT product categories should be licensed. For cars, the industry has clearly accepted end product-level licensing in practice (regardless of what some may say when lobbying politicians). For smartphones, it’s been the standard approach for a long time. But there is even one major company, Qualcomm, that considers the end product the right licensing level for cars and smartphones, yet believes in chipset-level licensing with respect to a variety of IoT products. Other companies argue that the licensing level should be the same across all product categories.
There are different approaches that are being tried. Some believe that the way to make cellular connectivity affordable in the IoT space is to give licensees only a subset of the standard and to limit applications to low-bandwidth and/or infrequent transmissions. Others believe it’s better to give licensees the full stack.
At this point, different industries are dealing with the topic and trying to work out solutions. If SEPs can be overleveraged too easily, IoT innovation may indeed suffer, but if regulation complicates SEP enforcement, wireless innovators won’t be fairly rewarded.
There are court rulings that one may or may not consider reasonable, but at least for now there is no hard evidence of a crisis that requires drastic and urgent intervention.
By coincidence, SEP licensing in various segments of IoT is in the process of being worked out while the Unified Patent Court (UPC) is soon going to hear and decide its first SEP cases. Decision makers could, and arguably should, take a step back instead of rushing to a decision the way the European Parliament did before the elections.
Nokia v. Verifone is actually an opportunity to obtain further clarification, as Nokia brought cases not only in national courts but also in the Munich and Mannheim Local Divisions of the UPC.