ABR demonstrates adaptive controller to Justin Trudeau
Blog post posted May 11, 2017 by Peter Suma
On May 11, 2017 ABR demonstrated ABR Reach, our neuromorphic adaptive robot arm controller, at the grand opening of OneEleven's new space.
OneEleven is Canada’s largest scale-up innovation hub, and the home of ABR’s Toronto office. Prime Minister Justin Trudeau attended the grand opening and visited with Applied Brain Research.
ABR is a leading edge neuromorphics company. We use our neural computing methods and tools, such as our neural compiler Nengo and ABR Reach, to develop leading edge solutions for customers in cognitive computing, robotics and large scale manufacturing markets. Whether it’s assembly processes that cannot be automated with traditional robotics or research into the feasibility of compliant robots designed to work with humans, ABR brings the best team of AI, robotic and neuromorphic scientists to work on our client’s challenging projects. ABR has a strong base of existing technologies, such as Reach, available for license to speed the development of customer solutions.
Dr. Eliasmith demonstrated the ABR Reach robotic arm controller interactively to Prime Minister Trudeau, highlighting Reach’s ability to adapt to unexpected forces and unknown tools, just like the human motor control system after which it was modelled.
ABR Reach began as the PhD thesis of Dr. Travis DeWolf, under the supervision of Dr. Eliasmith at the University of Waterloo’s Centre for Theoretical Neuroscience, and has since been extended and patented by ABR. The original research was published in November, 2016, and Reach is now being advanced to increase its tool manipulation abilities for use in commercial applications.
For the demonstration ABR used Reach to control the Kinova Jaco2 robotic arm, a great Canadian robot whose Montreal headquarters the Prime Minister visited earlier in the year. The ABR team was impressed with the Prime Minister’s detailed questions, showing his general understanding of robotic control systems and disruptive potential and applications of artificial intelligence in the robotics markets.
You’re taking this to the next level with how it thinks, how it learns, and how it reacts…the simulated neural networks you’re creating through this can apply to this [robot] – it’s just a great visualization – but of course this applies to thousands of things.
To learn more about Reach and our work on adaptive control, see the adaptive control page.