Sim2Real AI Helps Robots Think Outside The Box

There is a paradox in AI and robotics: what’s hard for humans (such as playing chess) is easy for robots, and what’s easy for humans (such as picking up postal parcels) is hard for robots.  

This image has an empty alt attribute; its file name is AmbiSort_Adaptable-AI.gif

Consider items you interact with on a daily basis: a tube of toothpaste, a paper cup filled with hot coffee, a bag of chips or a bottle of water…

You know that these items may be partly occluded by other items, that they deform in different ways, and that some items are transparent. You handle these everyday without thinking, but these items can be challenging for robots.

At Ambi Robotics, our robotic systems learn how to handle diverse items using data generated by advanced simulation.

Using Simulation as a Robot Classroom

We fine-tune our simulations to the performance of our sensors, our robots, and variations on the items our robots will handle. Our simulations run extremely fast, hundreds of times faster than robots training in the physical world, so we can train our robots overnight. This is what enables our solutions to work reliably from day one.

There is always a gap between simulation and reality, known as the Sim2Real Gap. At Ambi Robotics, we narrow this gap by fine tuning our simulations as we encounter edge cases and collect more data.  We do this efficiently by staying at the cutting edge of research.  

Are You Up For The Challenge?

Overcoming challenges like this is what we do at Ambi Robotics. We believe in a future where robots work hard so humans work smart. Our approach to solving problems focuses on human-centric design that prioritizes people every step of the way, hardware-software codesign that exploits hardware and software synergy,  and collaborative innovation that unlocks true evolution.

Does that sound like the sort of environment where you’d like to be? We’re always looking for people who bring new ideas, inspiration, and energy to our team. 

Scoot on over to our careers page and let us know if you see an interesting fit. Perhaps the next evolutionary interaction in AI will be you and us.  

Yi Li, Head of Computer Vision at Ambi Robotics