Engineering an AI-powered autonomous driving system to automate the logistics industry from dock to door.
In May 2017, Google’s AI-powered AlphaGo beat Ke Jie, the world’s best Go player, because of one simple reason: rules. AI loves them. But what makes AI godlike on game boards, has yet to be successfully translated to the open road. AI systems cannot reliably navigate a vehicle in the real world. There are simply too many unpredictable variables; infinite unique situations, infinite rules. Truly reliable and safe autonomous driving requires a system with a flexible commonsense understanding of physics and the actions of others.
Enter ISEE. The startup is engineering next-generation, humanistic AI to automate the logistics industry from dock to door. ISEE was the first autonomous driving company to achieve exit-to-exit autonomous highway driving, the first to merge onto a highway in heavy snow, and can handle congested traffic better than one of the most well-funded autonomous driving startups in the world.
ISEE’s core technology is purpose-built for complex environments with high uncertainty (like shipping yards and congested highways). It understands context and infers intentions of other drivers to navigate all possibilities as they arise, unlike other AI solutions that require hard-coding. Co-founders Yibiao Zhao and Chris Baker based the technology on theory of mind, or the ability of humans to comprehend the intents and beliefs of others. They realized that if they could design an AI that could infer, then that AI could successfully navigate dynamic and unpredictable roadways. It wouldn’t need to know all possible solutions to a problem, it would just need to know how to react in the moment.
As noted in a 2017 interview with the MIT Technology Review, “‘When we see something for just a few seconds as we’re driving, we can quickly deduce the intention of the driver in front of us,’ said Zhao. ‘We want cars to have this same predictive capability.’
To give it to them, Zhao and his colleagues have drawn on some of their earlier research inspired by cognitive science. That included teaching a robot how to crack a nut using a hammer, and then taking the hammer away and getting the robot to select the next most appropriate tool from a random assortment.”
The logistics industry is not one-size-fits-all. And that’s precisely why ISEE chose to pursue it. With its multiple vehicle types, highly variable operational settings, and shortage of drivers, logistics is uniquely suited to the value ISEE can provide. Its AI can boost efficiency in a shipping yard just as easily as it can navigate a highway trip or the stop-and-go traffic in a city at rush hour. And for logistics companies, there’s no need to buy new vehicles—the entire ISEE platform can be retrofit onto existing infrastructure.
In just over two years, the team of scientists and engineers at ISEE has built a platform that behaves, in many ways, like a human, but with the reliability that only AI can provide. It is in the platform’s unique combination of humanistic reasoning and computational logic that will give the world the first truly safe autonomous vehicles for any industry in any driving situation.