Overview
With great power comes great responsibility. Recent advances in artificial intelligence and robotics have unlocked new capabilities and potential for autonomous self-driving vehicles. However, bringing these vehicles outside of the lab into the real world requires careful consideration of safety and reliability. Ryan Gariepy of Clearpath Robotics & OTTO Motors will not only tell you why starting by putting self-driving vehicles on the road as a first step isn’t a good idea, he will also tell you how his team is able to take advantage of state of the art research without compromising safety. Finally, he’ll answer the timeless question of ‘when is a car like a couch?’
Objective
To outline the challenges of deploying modern artificial intelligence techniques in production, and safety-critical environments as well as lay out ways to overcome these challenges.
Target Audience
Developers, designers, tech entrepreneurs, investors, and researchers
Five Things Audience Members Will Learn
- Why self-driving vehicles in factories are a much better bet than self-driving passenger vehicles
- The difference in risk between a ‘typical’ application of AI and using AI in self-driving vehicles
- An approach to applying modern developments in AI on multi-ton robots in a safe and reliable way
- What the differences are between the ‘state of the art’ method of designing full-stack robot control software and more traditional methods
- How modern techniques can fail in unpredictable ways
