IP and technology: Smart moves

In 2019, when it comes to making a home “smart”, the possibilities are endless. Since the introduction of Amazon’s digital voice assistant Alexa in 2014, the smart device landscape has expanded into all aspects of daily life.

A handful of examples include camera-enabled doorbells; the power to control your entire home’s energy system from a single device; and a fridge that can take a picture of its contents and send it to your smartphone when you’re shopping at the store.

With so many products entering the market, what can innovators do to make sure their inventions are adequately protected?

Patent problems

When seeking out IP in relation to smart devices, companies may struggle with the traditional route.

James Reed, senior patent counsel, says companies should first consider whether patenting a device can offer any meaningful protection. This is because smart devices operate using software and it can be difficult to know when a competitor’s algorithm is infringing your own.

"There’s no one player, which means there’s no one player to accuse of infringing." - Peter Finnie, Gill Jennings & Every

“With a patent you have to prove someone is infringing it, which may not be possible because the software and algorithm associated with an accused smart device operates entirely within the cloud and is inaccessible to a third party,” he says.

Matthew Howell, a partner at Haseltine Lake Kempner in Bristol, says this problem is heightened by the fact the patent landscape in this market has become extremely saturated.

“We’ve probably already entered the stage where you’re not going to get really broad, conceptual patents. More likely, you’re going to get narrower, incremental patents for minor implementation details which can be relatively straightforward for competitors to work around,” he says.

Additionally, Howell says, a program or algorithm can be written in a multiplicity of ways, so unless a company has a patent to cover them all, establishing infringement can be difficult.

Not so smart

Another problem when it comes to protecting smart devices is the distributed nature of the innovation.

“When you say ‘smart device’, it’s almost certainly not as smart as you think,” says Peter Finnie, a partner at Gill Jennings & Every in London.

“This is because most of what that device can do is not done on the device itself, but on a remote network or server in another location,” he adds.

Rather than being in a field of its own, a smart device’s components—a physical shell, a network that could be in a different country, and a platform operator—make it an amalgamation of IP.

This “pick and mix” aspect means that one smart device is probably the product of multiple rights owners, potentially operating across different countries. This, says Finnie, directly conflicts with patents, which are very territorial.

“When it comes to defining the invention in a patent, there isn’t a traditional model. There’s no one player, which means there’s no one player to accuse of infringing,” he explains.

Another drawback with protecting smart devices is that many of them incorporate machine learning.

A problem here, Finnie says, is that machine learning often depends on using data from third parties, which can create questions over ownership.

For example, in the scenario that one company’s machine learning algorithm learns from the data provided by another party, both parties must figure out the terms under which the data was given. Simply, who owns what?

Trade secrets

When it comes to smart devices, companies should consider that trade secrets may offer better protection than patents, Reed says.

One UK-based company, Doppel, uses a combination of the two to protect its innovation.

Doppel is a smart device which works by creating a silent vibration on the inside of a wearer’s wrist which resembles the “lub-dub” of a human heartbeat.

The Doppel wearable has different rhythms: slower rhythms which are calming, and faster rhythms which help wearers feel focused, says the company.

Jack Hooper, a co-founder of Doppel, says the device’s hardware is patented but the software that drives it is not.

Doppel made the decision to patent the hardware when the device was first shown to the public, but kept the software secret. This is partly because of the complexity of patenting software in the UK.

“Additionally, depending on the strength of the patent, it could be fairly easy for other parties to work around it,” Hooper explains.

“The patent is powerful because it protects the device from being reverse-engineered, but it’s more powerful in combination with our trade secret, which protects the software against being copied,” he adds.

Patent offices across the world pose a number of challenges for those looking to patent smart devices.

Smart devices: UK

At the UK Intellectual Property Office (UK IPO) and the European Intellectual Property Office, software as such is excluded from patentability.

This is potentially one of the biggest setbacks for innovators. Howell says one way a company can get around this exclusion is by proving that a technical effect arises from its software.

But Finnie says this is still riddled with problems because innovators and IP offices alike are struggling with the definition of “technical”.

“There is no standard. The definition of technology as it’s understood by innovators is somewhat separated from the courts and patent offices,” he adds.

Finnie says the law has a prejudice against non-technical applications of technical solutions created by smart devices.

“Smart devices that simply let us lead our lives a bit better are difficult to protect because the law is so antiquated,” he says.

He explains that in the eyes of the UK IPO, a smart device which tells you when to take your medicine is seen as non-technical, but if you have a smart device that can automatically cut off your car’s engine, that’s technical.

“The computer scientists making both of those devices have the same skillset and would argue that they are both technical,” he says.

Under the European Patent Office’s guidelines for examination, machine learning is listed under exclusions.

“The guidelines on machine learning and artificial intelligence dismissed the whole field as being, essentially, just a mathematical method,” Finnie says.

“The problem is that this doesn’t recognise that the programming of machine learning and the training of it are technical considerations, he adds.

Finnie says this is stopping innovation.

“Without being able to patent an innovation, investors who want exclusivity are not going to invest, which in turn  limits innovation,” he explains.

As it stands, the innovation is ahead of the law, but he predicts this will change in time.

“The law does evolve, it just lags,” he says.

Smart devices: US

Inventors in the US also face concerns about whether the smart device they have created will be eligible for patenting. The eligibility criteria for smart devices stems from the US Supreme Court decision in Alice v CLS Bank, when the court ruled that an abstract idea is not eligible for a patent simply when it implements that idea on a computer.

Reed says one of the biggest problems patent practitioners face is the continuing lack of consensus in the US court on what does, or does not, qualify as “abstract”.

There are arguments for and against Alice. On one hand, it limits some notorious non-practising entities’ business practice of amassing a large number of patents directed to flimsy concepts that reflect no legitimate, inventive effort commensurate with the grant of a patent right.

“If there was no requirement that an eligible patent claim is limited to the inventive, specific solution to a technical problem, our patent system would be awash with thousands of patents that claim nothing more than known computing technology applied to well-known business practices,” says Reed.

“However, Alice can also negatively impact innovation because it imposes an intractable standard for what can be patented.”

Reed says small companies making smart devices might find that a bigger player in the industry is infringing their product, but after Alice it can be so hard to get a patent that they may not be able to protect themselves against infringement.

Smart device: India

Shyamal Kishore, a manager at Sagacious IP in India, says the law there is more stringent than it is in the west.

While in the US there is prior case law, such as Alice, there is no clear-cut framework for what can or cannot qualify for a patent in India, he says.

“The law here gives us a sweeping statement that software cannot be patented and that’s it,” laments Kishore.

Therefore, when registering a smart device for a patent it can be very difficult to argue its patentability.

“It’s like a coin toss—when you argue your case before the examiners you might hit the bull’s eye, or you might not,” he says.

“The law is stringent not because it poses a lot of requirements, but because it lacks clarity,” he adds.

This is changing, however. In the last three or four years, the volume of software patents being filed at the Indian patent office is increasing.

Kishore says there is mounting pressure on the IP office, and the burden of proof is being passed down to the examiners to establish why a particular software patent cannot be passed. This lack of legal clarity is not, however, depressing innovation.

“As more and more patents are filed, eventually the patent office will have to clarify the law,” he concludes.

Doppel fact file