The most advanced robots have the ability to sense and interact with their environments. That’s why advances in sensors are one of the five megatrends we have identified that will transform robotics startups.
Historically, sensors were either too limited or too expensive for most applications. Ten years ago, lidars were confined to high-end robotics applications and military uses, making them expensive and difficult to integrate. Looking back to the 2005 DARPA Grand Challenge, Carnegie Mellon fielded two cars each with five lidars (one Riegl Q140i and four SICK LMS 291) and a radar (Navtech DS2000). For the Riegl, they also had to design and implement an advanced stabilization gimbal to allow it to collect valid data. These sensors alone cost well over $100,000, and the best of these lidars could only collect data at a rate of ~13,500 points per second.
Fast forward to today. Companies like Airware (Lemnos portfolio company), Cruise Automation, and Marble (Lemnos portfolio company) are greatly increasing the number of drones, self-driving cars, and robots in the field. As a result, the demand for advanced sensors, like lidars, is growing exponentially. In turn, this demand is leading to incredible advancements in capability and drops in price. For instance, Velodyne has a new lidar on the market for ~$8000 that can do 300,000 points per second.
But that’s just the beginning. As self-driving cars become standard, the price drops will be incredible. In a recent Forbes article, Velodyne “says it has designed a new solid-state version of its technology that provides 3D imaging for automated vehicle systems that will cost less than $50 per unit when manufactured at high volume.”
Why does this matter for startups?
At Lemnos, we talk about how we like to invest in companies that are “surfing Moore’s law”—companies founded today with unit economics that don’t work but are in an industry with a high likelihood that the unit economics will change. Right now, if your robotics startup needs a few $8000 lidars for your robot, your business may not work. However, if the price of those lidars dropped to $500-$1000 a piece, your business may succeed. In which case, now is the time to start that company.
Currently, lidars can be 50 percent of the cost of a robot, so even a modest reduction in cost can make a business viable. Venture capital will happily help give you a head start on your competitors to develop the rest of the technology and business (which will likely take a year or two anyway).
This isn’t just true for lidars. Another technology that has seen improved capacity and dramatic price drops is high-precision GPS. Our portfolio company Swift Navigation has been a pioneer in ultra low-cost high-precision GPS. Historically, these units were at least 10 times more expensive than what is now available.
Inertial measurement units, radars, encoders, and other transformative technologies are all poised to hit the same type of demand curve. There are probably others that we haven’t thought of yet. If you want to share your ideas with me, find me on twitter @nomadicnerd.