Predicting the World Around Autonomous Vehicles: Our Investment in Perceptive Automata
As I’ve said before, cars are “social.” They exist alongside other human-operated vehicles, cyclists, and pedestrians. When we’re behind the wheel, we constantly survey the roads looking for clues to help predict what other people will do. Will that teenage skateboarder jaywalk? Will the minivan driver speed up as I try to make an unprotected left-hand turn? Who goes first at a four-way stop if we all arrive at the same time?
People use a “theory of mind” to face those kind of split-second decisions all of the time. However, what comes relatively easily to us humans is incredibly difficult for autonomous vehicles. To improve safety for passengers and pedestrians alike, it is so important to have an intuitive self-driving system that is able to recognize, understand, and predict human behavior.
Cracking that middle layer of the perception-prediction-planning stack is exactly the challenge that Perceptive Automata is tackling. We’ve been intrigued by Perceptive’s work since our initial meeting soon after their founding, and I’m thrilled to welcome the Somerville, Mass.- and Sunnyvale, Calif. based company to the Toyota AI Ventures portfolio.
Although I can’t divulge too much about Perceptive’s technology, they use behavioral science techniques to characterize the way human drivers understand the state-of-mind of other humans and then train deep learning models to acquire that human ability. These deep learning models are designed for integration into autonomous driving stacks and next-generation driver assistance systems, sandwiched between the perception and planning layers.
These deep learning, predictive models provide real-time information on the intention, awareness, and other state-of-mind attributes of pedestrians, cyclists and other motorists. Here’s a video of their product in action:
Perceptive’s founders CEO Sid Misra and CTO Sam Anthony met in the Vision Lab at Harvard. Together, they are a powerhouse founding team with the academic credentials and hyper-focused ambition to help machines understand and predict human behavior. What’s more, Sam and Sid are surrounded by a stand-out team of data scientists, machine learning engineers, and researchers.
We are excited to back this impressive team in their $16M Series A round, alongside JAZZ Venture Partners, Hyundai CRADLE, First Round Capital, and Slow Ventures, among others. As for the road ahead, the team at Perceptive is looking to recruit more talent that’ll be needed to grow their product and deployment teams. If you’re passionate about building a safer future, they’re hiring!