Wevolver and Syntiant are creating a series that explore the work of innovators and the future of pervasive AI. Syntiant is developing ultra-low-power AI processors. Because they believe in the importance of innovation, Syntiant is engaging in these fireside chats with engineers and designers who are on the cutting edge of their field.
In previous conversations we spoke with CEO and founder of Women in Voice Dr. Joan Palmiter Bajorek, The Things Industries CEO Wienke Giezeman, Rev.com’s Daniel Kokotov, Star Wars Animatronic Designer Gustav Hoegen, and Fashiontech Designer Anouk Wipprecht.
Tiny Machine Learning- a fusion of ML techniques, algorithms, and hardware that allow you to process data on ultra-low-power devices on the edge of a network, packs a punch. Getting real-time insights from frugal models could unlock better sustainability from littering sensors throughout intelligent cities and environments. Moving from big to little is not without teething problems, but the diversity of the subfield is what Gousev believes breeds the most promise and excitement. “You give this technology to people, and they discover ways to use it for different applications. If you look at verticals, I see a lot of traction in tinyML and many opportunities in industrial IoT, because all industries are going through digitalization.”
Gousev, who leads a Research and Development team focused on hardware systems, notes predictive maintenance, consumer electronics, and smart cities as proliferating use cases enabled by tinyML. One area he’s particularly interested in is accelerating healthcare.
“Healthcare is usually a bit behind because of the regulations, data, and privacy. However, you can accomplish a lot in the space by putting tinyML devices on your body or in your environment, then and use this actionable information for whatever you're trying to discern. For example, many companies are developing better hearing aids using tinyML technology. There are boundless opportunities in this space – you can use as many examples as and as wild your imagination is.”
The hype curve that accompanied the explosion of deep learning led to a bit of gluttony, with the tantalizing temptation to throw the technology at every new problem. Overuse can have byproducts of unnecessary power consumption and useless insights.
“You can use deep learning, but it’s not a panacea. It’s all about problem-solving, which tinyML enables on the technology and mindset side. It’s not about brute force approaches. It’s extremely important for people to realize that it’s not only deep learning. There are other tools. What we have seen recently in the past year or so is that many companies are adding classical algorithms to their tool chain, for example TensorFlow and classical algorithms, MATLAB is adding classical algorithms it in addition to CMN type of algorithms, because people realize you need to give these tools to the people. Then they will figure out what works best.”
Gousev co-founded the tinyML foundation, bringing slews of intelligent brains together to work on tinyML for good to democratize the technology. His vision to empower users - not just subject matter experts - led to the birth of the foundation, which was started at Qualcomm roughly ten years ago. At its genesis, Gousev was working with partners like Google and also Syntiant, who was responsible for pioneering low-power research, and many other companies, to build up the ecosystem.
“The beauty of the community is also in its diversity – you can do a lot of things in one company, two companies, and five companies but today, we have 45 tinyML groups in 37 countries, basically covering the whole world. Different countries have different aspects and different priorities in tinyML. For example, companies in Africa are developing these technologies to help their communities in various ways. Really, the community is very important in terms of the ideas, the brain power but also diversity. Everyone is bringing their unique end of the table. The tinyML Foundation’s mission is to see a new world with a trillion devices enabled by tinyML that work together autonomously to make a better and more sustainable world for all of us.”