We were lucky to catch some time with Gene Vinokur, who regaled us with updates on the most exciting technology currently under MERL’s belt. We also discussed MERL’s patent strategies, trends in the patent world, and how futuristic technology could change our lives.
Leaders League. Could you talk us through some of the most exciting technology MERL is developing?
Gene Vinokur. We have over 60 researchers in various fields – pretty much everything except pharma, biotech and chemistry. We’re active in applied physics, AI, robotics, data analytics, autonomous driving, signal processing, speech transcription, optical communications, power grids... It’s easier to talk about what we’re not doing! Our research comprises two parts. There’s our applied research, which we do for Mitsubishi Electric, looking at problems we might face in the near future, perhaps three to five years from now.
Then there’s our fundamental research, looking at trends, developments and technology subjects extending to five to ten years in the future. Both arms aim to create new knowledge and furthering the state of the art. Wireless energy transfer is exciting. So are big data, autonomous driving, and additive manufacturing in zero-gravity. Optical communications are also exciting – we [in a joint project with University College London] were the first to transmit 1TB of data per second. It’s hard to pick out something specific!
How do you decide what IP is most important to protect, and how do you form strategies to protect it?
We use patent applications and claims to create strategies for how to protect and utilize the knowledge created by our research. This not only protects the results of our research, but demonstrates to our parent company how this research can benefit their business. “Protect and promote” is our mantra. Patent protection isn’t just a legal or technical document: it’s a marketing document.
In that regard, our patent applications are very different from scientific papers. The latter share knowledge; the former protect practical applications of that knowledge, also to change society for the better. To do this, we’ve developed our own in-house framework. It has four layers: firstly, a researcher provides a summary of his/her findings. Then there’s the attorney summary: a strategy of how to protect this research, sometimes different to what the researcher might expect. Then the invention is described in more detail according to that strategy. Then the patent application is finalized. The second stage, the attorney summary, is very important: it’s where claims come in.
We’re storytellers – that’s important! The story in a patent application needs to showcase the inventor’s contribution to the art, as well as explain why he/she deserves the patent. This helps during prosecution, and to promote our research to our parent organization. It’s also translates more easily into different languages, such as Japanese. Sometimes, the link between the knowledge and its practical application is not immediately obvious. But our job is to clarify it.
Will MERL be making much use of blockchain or quantum computing? If so, how?
We’re doing some research in this regard – not extensively, but to some degree. For blockchain, our research is still active so I can’t talk too much about it, but there are two stages: the protocol itself, i.e. how blockchain is implemented, and how that protocol is used. We’re doing research in both directions: modifying the protocol, and working on some practical patent applications that can take advantage of the modified product. I am convinced that blockchain technology must have technology for some applications. It’s a matter of convenience. I don’t see anything there that can’t be implemented in the most centralized manner using other technologies. But I can understand that in some situations, when a lot of different unrelated people are involved, blockchain computations can be advantageous. Sometimes we’re thinking about using blockchain to do social good, to make life better organized for many people.
"Business needs to trust the research and its viability and reliability. To build trust, our researchers have to draft patent applications that explain the rationale behind those patents"
Quantum computing is exciting, but there are limitations. You need a repository of atoms and electrons in an extremely cold place, close to absolute zero, otherwise the environment is too noisy to make subatomic measurements. So we’re talking about mainframe computers built by IBM or Google, in huge facilities deep underground, and because it’s so cold you need different hardware as well as different software to handle the quantum elements.
We’re working on how we can implement logic in such a cold environment, and how to increase the temperature of the environment, even by five Kelvin, so it’s a less hostile environment for measuring. But there’s a lot of work to be done. Cloud computing, in combination with AI and quantum computing, could be very useful. But some researchers are skeptical, because of the cold temperature requirement.
How could quantum computing be useful?
Secure communications is the most obvious one. But in our world right now, everything is about data. Your question is as broad as 'Where can the internet lead to?'. Everything is data-related now – whatever we see, touch, do, including our politics. We generate, process, deliver and transform data. If quantum computing makes the processing of big data instantaneous, it could lead to unimaginable things.
This plus AI plus energy sources could change our lives like the internet has. I’m not sure about this talk of AI simulating human brains, but it can help us greatly in virtually any number of ways. The IoT is one current example of working with big data.
What developments in innovation were getting high publicity a few years ago, but have since failed or undergone slow growth?
This is hard to point out. Our products develop along different timelines to the business itself – so maybe a given technology didn’t pick up when commercially needed, but could be a hot commodity ten years later.
One example involves adaptive control, which is the control method for when you’re trying to control a system you’re not very sure about – let’s say, a train, where you don’t know the weight of the train or how much friction is on the track. We have many patents protecting different adaptive control implementations.
But the problem here is trust. Business needs to trust the research and its viability and reliability. Just because it seems to be working on your computer most of the time doesn’t guarantee that it’ll work all the time, which can be problematic. It’s like autonomous driving: it works most of the time, but from time to time it kills someone, and we don’t want that. So to build trust, our researchers have to draft patent applications that explain the rationale behind those patents. We can deliver what we’re promising to deliver. But convincing the business heads is an uphill battle.
What predictions do you have for the next five years in innovation?
We need three components: energy sources, the computer power to process big data, and the means to deliver the results of that processing. We’re working on all three areas, and I think we’ll make great progress. This technology will work towards globalization and unification, and against global warming. We need to learn how to live together.