Welcoming Applied Intuition to the M12 portfolio
The venture capitalist’s job is to identify the rare combination of unique technology, unsolved problem, and world-class team, and provide such a team with the capital they need to achieve their vision. M12, as the three year-old venture firm backed by Microsoft, has significant experience here, deploying hundreds of millions of dollars into startups spanning the (physical and topical) globe — from teams working on issues of fraud in London, to 3D printing innovating materials in Boston, to writing more secure code in Israel. Of course, the topic of autonomous vehicles has been on our minds since inception. We’ve been looking for that rare combination in this segment but also with software as its point of leverage.
When we first met Qasar Younis and his team in 2018 we quickly saw they were doing something fundamentally different — bringing established best practices in software development and testing methodologies to A/V testing, easing the burden on engineers to scale the building of safe and reliable systems. Every year, tens of thousands of people lose their lives and millions are injured in vehicle crashes in the United States alone. The vast majority of those crashes are caused by humans. While technology has the potential to be better than humans, the question of how to test and demonstrate that improvement (without driving billions upon billions of “test miles”) is a major challenge. Borrowing concepts from the EDA market (tools that help design and simulate the performance of semiconductors), Applied Intuition aims to make autonomous systems development more efficient with simulation playing a key role.
Whilst environmental perception is not a solved problem in autonomy development, motion and path planning are considered to be the greater challenge in building autonomous systems which will behave reliably and predictably in the rich diversity of the real world. Instead of optimizing just for visually stunning 3D worlds that are appealing to the eye, Applied Intuition has invested heavily in helping A/V engineers understand the algorithmic performance of an autonomous robot as it perceives and acts on its environment. When combined with how the company’s products prepare data extracted from real world driving miles, doing parameter sweeps to automate creation of test scenarios including synthetic variants, easing regression testing, a powerful IDE for A/V engineers is at hand. And their customers love it.