If you want to catch a glimpse of venture capitalists in their natural habitat, check out the robotics meetups around town. That’s a takeaway from one such gathering of researchers and demos last Wednesday, hosted by the Wyss Institute and Harvard’s School of Engineering and Applied Sciences.
A quick look around the room and you’d find investors from Google Ventures, General Catalyst, and Matrix Partners (to name a few) sniffing around for promising technologies and entrepreneurs. Most of the projects on display were too early for VC funding, but it’s always better to know what—and who—is coming down the pike before hearing a pitch.
Another takeaway: Harvard has been making a real push on technology licensing and startup creation in areas like computer science, artificial intelligence, and robotics. The Wyss Institute in particular, founded in 2009, has assembled a large and ambitious staff on the technical and commercialization sides, to help push its research in biologically inspired engineering into the real world.
Three quick highlights from the meetup, in rough order of commercialization progress (definitely not meant to be comprehensive):
1. RightHand Robotics
This is a startup company formed out of Robert Howe’s longstanding Biorobotics Lab at Harvard and the Yale Grab Lab. Co-founded by Yaro Tenzer, Leif Jentoft, and Lael Odhner, the company has developed a three-fingered dexterous gripper that can pick up (and put down) a variety of items such as tools, produce, and dry goods. The key is to do it accurately, cheaply, and without breaking anything. RightHand Robotics is looking at applications in order fulfillment and distribution; it could find customers in fields like automotive, pharmaceutical, healthcare, and food packing.
In the photo, the robot hand is picking up a ball using tactile sensors only—it’s feeling its way through the task, and making mistakes along the way. This particular demo is a research project by Harvard graduate student Qian Wan. The RightHand gripper is attached to an arm made by Universal Robots (recently acquired by Teradyne).
RightHand is playing in the same general ballpark as Rethink Robotics, Kiva Systems (owned by Amazon), Harvest Automation, and Fetch Robotics.
2. Soft Exosuits
One of the most popular stations at the meetup showed off DARPA-backed research from the Harvard Biodesign Lab, led by Conor Walsh. Several “soft exosuits” were on display—these are wearable-robot contraptions that could help boost people’s ability to walk for long distances, or help stroke patients regain strength in their arms or legs. The basic idea is to provide power to a person’s hip or ankle joint, say, using wearable sensors, actuators, and a flexible apparatus that conforms to the body’s natural mechanics.
Easier said than done: researchers have been working on powered “exoskeletons” for at least 20 years. I got some details from Shane Stilson, a Wyss Institute engineer who came from Utah robotics company Sarcos, by way of its acquisition by Raytheon in 2007. It sounds like the softer, lighter aspect of the exosuits makes it so people might actually feel comfortable wearing them; the new designs may also consume less power than past devices. There’s no startup company here as of yet, but technologies are already being licensed.
3. Robot Swarms
Radhika Nagpal’s group at the Wyss Institute studies what it calls “self-organizing collective systems.” Think the behavior of a flock of birds applied to robots, and you’ve got the right idea.
The “Kilobot” project is apparently the largest-scale demonstration of robotic swarm intelligence to date. More than a thousand little robots—each is a quarter-sized cylinder with three vibrating legs—can assemble themselves to form any two-dimensional shape (like a star) commanded from above. Each robot communicates via infrared and acts on local information from neighbors around it. The process can take several hours, but it shows how complex organizational behaviors can emerge from lots of relatively simple nodes in a network.
So far, Kilobot is mainly intriguing science, with possible applications in search and rescue, robotic error correction, and collective artificial intelligence. No word on whether it can explain the mentality of Silicon Valley entrepreneurs or VCs—but it may eventually attract them.
