Guest post by Kenie Ho and Charles Huang
You might be riding to work in a driverless car without ever having to look up from your text messages. Or you might rely on weather forecasts derived from micro-weather patterns using the barometric sensor of every iPad inside a local area. These kinds of IoT miracles will use dozens or even thousands of IoT devices. That creates challenges for a company trying to protect its IoT innovations, or for a company trying to avoid infringing someone else’s protected technology.
Businesses typically protect their R&D through patents. A patent allows an inventor to exclude others from making, using, selling, or importing a patented invention.
Companies, like IBM, Intel, and Qualcomm, recognize patents as potent business tools because they can use them to keep competitors out of a market or obtain lucrative licensing royalties by allowing the competitors to practice the inventions. In 2015 alone, these companies each applied for and obtained several thousand new U.S. patents, many on IoT-related inventions. Experts estimate that more than 20,000 patents and patent applications covering IoT technologies exist world-wide.
But are IoT patents truly valuable?
I Want to Sue You, But I Can’t
At its core, IoT is basically a massively distributed network. IoT devices across the world work together to implement creative IoT solutions. Because of that, it can be challenging to obtain patents that are useful against competitors in the IoT space.
To prove infringement in the United States, a patent owner must show a single entity infringes the invention claimed in the patent. In an IoT ecosystem where many devices and actors must interact to implement a use case, it can be difficult to meet that requirement. Often times, no single entity implements or uses the entire claimed invention.
For example, in the autonomous-driving scenario, smart cars might have IoT-enabled sensors reporting on the vehicle’s position, nearby obstacles, speed, and vehicle status. IoT sensors embedded on a smart highway, in a smart city’s traffic-control system, and around a smart parking lot might provide additional information for routing and safety. These devices might communicate with each other to pick up commuters, drive them to work, drop them off, and then park their cars—all without them lifting a finger.
The problem is that different entities own or manufacture each component in this scenario. And if a patent owner has a patent covering this situation, who can it sue for infringement? The smart-car manufacturer? The county maintaining the smart highway? The city with its IoT-enabled traffic system? The owners of the smart parking lot? Depending on how the patent was prepared, the patent owner might be able to sue some, none, or all of them.
If you have never read a patent, consider yourself lucky. It is an arcane combination of technical writing and legalese that will put all but the most stalwart patent attorney to sleep. And the most arcane section of the patent—called the “claims”—happens to be the most important because it describes what the inventor is actually claiming as the invention.
Under U.S. law, an entity infringes a patent only if it practices or uses the invention described in the claims. Without getting into all of the legal details and numerous exceptions, if an entity practices or uses only a portion of the invention described in the claims, it is typically not liable for infringement.
In the autonomous-driving case, if a patent claims the combination of using a smart car, a smart highway, a smart traffic-control system, and a smart parking lot, then an entity that practices or uses all of them in combination is liable for infringement. But if there are multiple entities acting in concert, and each practices or uses only a part of the claimed combination, then a “divided infringement” situation exists and the patent owner might not be able to sue any of them for infringement.
For these and other reasons, patent attorneys consider it a best practice to procure patents with claims targeting the actions of individual entities. A well-designed portfolio of patents might include (1) a patent directed to the smart car made by the manufacturer, (2) a patent on the smart highway maintained by the county, (3) a patent on the smart traffic-control system owned by the city, and (4) a patent on the smart parking lot run by the parking company. The patent owner would then have a portfolio of patents to choose from when deciding whom to sue (e.g., the car manufacturer, county government, city, or parking company, respectively)—preferably the entity with the deepest pockets.
But what happens if the novelty in the invention comes from the combination of all the “smart” elements, and the patent office will issue only a patent claiming the combination? Enforcing this kind of patent in a divided-infringement situation is much harder, but still possible.
In 2015, the U.S. Court of Appeals for the Federal Circuit—the highest court overseeing U.S. patent cases besides the U.S. Supreme Court—explained that an entity can still be held liable for patent infringement if it controls or directs multiple entities to jointly use a patented invention. That is, an entity would be liable for divided infringement if the acts of the other entities can be attributed to the first entity.
For example, if a smart-car manufacturer has a contractual relationship obligating other entities to embed and use IoT sensors on the highway, in the traffic-control system, and around the parking lot to implement the autonomous-driving use case, then the smart-car manufacturer could be found liable for infringing a patent claiming the combination. But unless the patent owner can show this type of control or joint enterprise, it will likely not be able to prove infringement for that combination patent.
Besides divided infringement, another obstacle facing IoT patents is territorial scope. A U.S. patent grants rights in the United States. Thus, a U.S. patent presumptively does not confer any protection to infringing acts outside of the United States. This poses a problem for IoT patents because many IoT use cases employ devices located outside of the United States.
For instance, in the autonomous-driving scenario, sensor data from a smart car might be routed to a server located in Canada—because it might be cheaper there—for routing and map updates before being sent back to the car. A U.S. patent claiming a “process” for autonomous driving that includes routing and updating maps would generally not be enforceable here because those routing and updating steps take place outside of the United States. But due to patent policy set by the government and U.S. courts, a U.S. patent claiming an autonomous driving “system” might be enforceable if the Canadian server was being controlled in the United States. The differences between the policy reasons for the two are beyond the scope of this article. The point is that territorial scope of a patent matters, particularly for IoT applications.
Good Patents, Big Consequences
A good patent that avoids the above problems and covers a competitor’s IoT products provides a big competitive advantage, especially if the competitor cannot design around the patent. Further, if the competitor had full knowledge of its infringing activities and had no reason to doubt the patent’s validity, but nonetheless continued with its infringing activity, it may be liable for willfully infringing the patent, an act that can triple the amount of actual damages.
The U.S. Supreme Court recently changed the law to make willful infringement easier to prove. Before the change, a patent owner needed to show, by clear and convincing evidence, that the accused infringer was reckless in infringing the patent and knew or should have known its infringing actions were reckless. Now, the patent owner needs to show by a preponderance of evidence—a lower standard—only that the infringement was “egregious” and not just simply a “garden-variety” infringement case.
In the past, if a company became concerned about a patent, it would seek a patent attorney’s opinion on the matter to avoid liability based on willful infringement. That practice went out of favor in the mid-2000s after the courts raised the standard for proving willful infringement. Now, with the lowering of the standard, that practice has enjoyed a revival if only to show that the company took due care in investigating the matter to reduce the likelihood of willful infringement and treble damages.
Despite the divided-infringement and territorial-scope issues, thousands of patents on IoT-related technologies are being issued each year. The key is to make sure to get patents that are well thought out to avoid divided-infringement and territorial-scope issues in the first place.
On average, it takes over 2 years to obtain a patent and most patents have a term of 20 years. It might be 10 or 15 years before the patent owner asserts the patent. How the market uses the invention can change significantly during that time. Thus, a patent applicant must carefully predict and anticipate likely infringement scenarios when protecting its IoT technology.
Kenie Ho has litigated over 50 patents in U.S. courts on electrical and consumer-electronics technology. He is a thought leader on intellectual-property issues for IoT and leads the IoT Legal Group at Finnegan, Henderson, Farabow, Garrett & Dunner, LLP.
Charles Huang prepares patent applications for IoT patents. His practice includes litigation, client-counseling, patent portfolio management, and patent analysis.
Patent Photo Credit to Nick Normal via Flickr.