Massive and rapid developments in both hardware and software are paving the way for the so-called ‘silicon brain’ (also see ‘neuromorphic computing’), which, in theory, would be able to ‘think’ and ‘feel’. Many believe that the silicon brain will play a pivotal role in enhancing machine intelligence to the point where it can solve some of today’s biggest problems and challenges.
There’s a long way to go before we may reach such a milestone, however. In the meantime, UK-based firm Opteran—an eight-year-old spinout of the University of Sheffield, led by Professor James Marshall and Dr Alex Cope of the same institution—has begun developing lightweight, low-cost ‘brains’ inspired by insects. And they’ve just last year raised £2.1 million in a funding round led by technology investment company IQ Capital to further their efforts.
The funding round also included Episode1, Seraphim Capital, various angel investors, and a Connecting Capability Fund (CCF) grant (a part of the Northern Triangle Initiative).
Artificial ‘Brains’ Inspired by Insects
For eight years, Opteran has researched insect brains as part of the ‘Green Brain’ and ‘Brains on Board’ projects.
Although insects have tiny brains, they’re actually capable of sophisticated decision making and navigation. This is enabled by a natural phenomenon known as ‘optic flow’, which allows insects to perceive depth and distance. By reverse-engineering insect brains, Opteran claims that it has been able to produce algorithms that require zero pre-training—allowing its technology to potentially mimic tasks like seeing, sensing objects, navigating, and avoiding obstacles.
Opteran will use its funding to pioneer ‘natural intelligence’, particularly its insect vision-inspired approach to low-cost, silicon-based autonomy. Pictured is a close-up of one Opteran’s points of inspiration: an insect’s compound eyes. Image Credit: Opteran Technology.
According to Opteran, its insect vision-inspired approach is an efficient, robust, and transparent way to achieve autonomy when compared to current deep learning techniques. In a recent trial, Opteran was able to control a very light (sub-250-gram) drone with complete onboard autonomy using less than 10,000 pixels from a low-resolution camera.
Solving ‘Fundamental Shortcomings’ in Autonomous Applications
Throughout 2021 and the months that follow it, Opteran says that the funding will be used to build out functionality in the algorithms and chipsets. This will include launching its following technologies: ‘Opteran Sense’, for collision avoidance and navigation; ‘Opteran Direct’, for SLAM (simultaneous localisation and mapping); ‘Opteran Decide’, for autonomous decision-making; and ‘Opteran See’, namely a 360-degree camera.
Said David Rajan, CEO of Opteran: “2021 will be the year when natural intelligence will challenge deep learning in solving some of the most fundamental shortcomings in autonomous applications, and this funding round will set Opteran on a path to be at the forefront of this next wave”.
The Future Potential of Opteran’s Technology
Rajan says that he is confident that, because Opteran is already in a position to demonstrate the technology, natural intelligence will become a highly sought-after way to deliver lightweight, low-cost, and effective autonomy in a new way that will bring growth opportunities for the robotics industry.
On top of this, the company’s Opteran Development Kit (or ODK) will allow engineers to apply the optic flow-inspired technology to a variety of applications in the robotics market. It’s hoped that the technology will eventually transform autonomous vehicles, drones, mining robots, and other vital use cases.
For more information on natural intelligence, read ‘Robots with insect brains’, published in Science.