All Posts

Last update on .

Realtime Robotics, Co-Founded By George Konidaris, Secures $2M In Funding

Click the link that follows for more Brown CS content about George Konidaris.

Imagine a self-driving car zooming down the highway, or a robotic arm welding car parts in a factory. Suddenly, in less than a second, another car swerves in front of it, or a human trips and falls into the work area. The difference between a close shave and catastrophe will depend largely on one thing: the robot's reaction time. 

In 2016, Professor George Konidaris (then at Duke University) of Brown University's Department of Computer Science and Professor Dan Sorin of Duke University founded Realtime Robotics, focused on developing on a processor that would allow robots to perform motion-planning tasks at up to 10,000 times faster than previous speeds. Last month, the company secured $2 million in seed funding to further its growth. Key investors in the seed round include SPARX Group Ltd., Scrum Ventures, and Toyota AI Ventures, a venture capital subsidiary of Toyota Research Institute (TRI).

Realtime Robotics enables complex robotic motion planning tasks to be accomplished up to 10,000 times faster than previously possible using a proprietary special-purpose processor, allowing robotic systems to instantly react to their environments and compute how and where to move as their situation is changing. This groundbreaking ability to instantly plan motion in response to rapidly changing conditions overcomes one of the primary challenges preventing robots and autonomous vehicles from achieving their enormous potential.

“Realtime has set itself apart by providing a novel solution with far-ranging implications” says Tak Miyata, General Partner at Scrum Ventures.

Uses for Realtime's lightning-fast processor are wide-ranging: it enables robots with sophisticated arms to be utilized in dynamic environments, dramatically increasing the types of industrial tasks they can perform. It can also be used by autonomous vehicles to help them operate at normal speeds —like humans, but safer— instead of slowing to a crawl when there is uncertainty regarding other cars, bikes, or pedestrians.

"I'm very excited about the technology we're developing at Realtime," says George. "When I was young, programming video games was much harder, because an incredible amount of effort had to go into generating realistic graphics on quite slow hardware. GPUs changed all of that, and led to a massive burst of creativity and excitement in computing gaming. The processor Realtime is developing will do that for motion planning. At the moment it's slow and difficult to get right, so there are virtually no deployed applications. Once we're done, the processor will open up a whole new world of possibilities for robot automation."

For more information, click the link that follows to contact Brown CS Communication Outreach Specialist Jesse C. Polhemus.