2019 will be the year of commercial class legged robots. That was the message delivered by Agility Robotics and Boston Dynamics during their respective opening and closing keynotes at the inaugural Robotics Summit & Showcase, produced by The Robot Report and WTWH Media in Boston.
Agility Robotics CEO and co-founder Damion Shelton updated attendees on its Cassie bipedal robot. Boston Dynamics co-founder and CEO Marc Raibert quickly discussed the wheel-leg hybrid robot Handle, which he said we’ll hear more about in 2019 with a real application, while focusing more on the Atlas bipedal and SpotMini quadruped robots. Raibert conducted a live demo of SpotMini (watch below) where the robot traversed a small obstacle and picked up a soda can and handed it to Raibert.
Neither company claims legged robots are a fit for every application. “If we evolved with wheels, I’m sure our environments would be good for wheels, too,” Shelton said. Raibert and Shelton both described potential applications for their robots, including construction, delivery, disaster relief and surveillance, but the availability of commercial class legged robot platforms to build upon will lead to innovative ideas.
Raibert showed video of Spot, the older version of SpotMini, being tested for home deliveries. He said Spot was sent to the homes of 20 employees, with the consent of their significant others, of course, and they had about 70-80 percent success rate.
Attendees pack the room to hear Agility Robotics’ CEO Damion Shelton deliver the opening keynote at the Robotics Summit & Showcase. (Credit: Derek Little/WTWH Media)
Legged robots gearing up for production
Cassie is gearing up for its second production run in July 2018, while SpotMini is in pre-production preparing for commercial availability in 2019. SpotMini will be the first robot Boston Dynamics commercialized in its 26-year history.
Raibert would not disclose the price of SpotMini. He said the latest prototype costs 10 times less to build than the previous iteration. Boston Dynamics is working with contract manufacturers to build 100 SpotMinis over the next year. It hopes to build 1,000 each year going forward.
Agility Robotics recently raised $8 million in Series A funding led by Playground Global, which was founded by Android co-creator and ex-Google robotics head Andy Rubin. The company is hiring mechanical, electrical, and controls engineers at its Oregon headquarters. It’s also adding employees for perception, business development and apps engineering at a facility in Pittsburgh.
Legged robots have long been challenging and expensive. The prices are finally coming down, but challenges such as agility, control laws, emergency stop, power consumption and stability will persist. Getting these platforms out into the masses is the only way to expedite their development.
The Dynamic Legged Locomotion Lab at the University of Michigan, for example, recently had Cassie riding a Segway to test custom controls. The University of British Columbia used deep reinforcement learning in simulation to test feedback control. Shelton said Agility Robotics is just starting to explore how deep learning can help bipeds.
Thirsty? #SpotMini will take care of that for you! @BostonDynamics @WTWH_Media @DesignWorld @therobotreport @SteveCrowe #RoboticsSummit2018 #robotics #robot #Engineering #Engineers pic.twitter.com/COxAinD6MH
— Paul J. Heney (@DW_Editor) May 24, 2018
Selling SpotMini as a platform
SpotMini will be sold as a hardware and software platform. The robot is flat on top with mounting plates for companies to hook into. SpotMini has a network connection and API so third-party software can talk with its software to develop apps. Boston dynamics is building its own apps, including a surveillance package, that it’ll use as reference designs going forward. There’s an optional manipulator arm sold separately.
Raibert said Boston Dynamics will conduct tests with potential clients in 2018 before rolling it out next year. SpotMini, which weighs 60 lbs., was scaled with the idea of making it small enough to fit in a house, an office or warehouse – environments that are really constrained.
Raibert showed video of SpotMini autonomously navigating, which is a new feature. SpotMini was manually driven through the space to create a map of the space using visual data from its cameras. During the autonomous run, SpotMini uses data from the cameras to localize itself in the map and to detect and avoid obstacles. Once the operator presses ‘GO’ at the beginning of the video, the robot is on its own.
All-Electric SpotMini solves Persian rug problem
“When we developed BigDog, which is bigger, we thought that having it roam around in the woods was the toughest mobility challenge that there was. The terrain is totally organic and unstructured,” Raibert said. “But it turns out that if you take a robot inside of the home. It’s a much tougher problem. Everything is really tight and cramped, depending on the scale of your robot.”
Raibert said one issue with legged robots negotiating tight spaces is they can’t scuff floors, walls or stairs. “You can’t go bouncing around. There’s a lot more geometry in the space that’s involved with negotiating.”
Part of the issue, too, is solving the “Persian rug” problem. “If you drop one drop of greasy oil on the Persian rug, you’re in trouble.” This is why SpotMini is all-electric and there are cameras on all sides to help it navigate tight areas.
Boston Dynamics reconsidering defense work
Skeptics have often wondered what has taken Boston Dynamics so long to commercialize a robot. And it appears the company’s mentality has changed since the Softbank acquisition in 2017. Raibert said Boston Dynamics’ long-term goal has been solving “the hard problems in robotics, leading to major new functionality.” Its new goal is developing products for real-world applications.
Asked if Boston Dynamics is considering defense work again now that it’s not connected to Google, Raibert simply said the company hasn’t decided. According to the AP, a federal contracting database lists more than $150 million in defense funding to Boston Dynamics since 1994. But that funding dried up in 2013 when Boston Dynamics was acquired by Google, which made it clear it wanted no part of in defense work.
Raibert said it’s a challenge balancing both the short- and long-term goals simultaneously. But he isn’t shy about asking for help. “I hope half of you quit your jobs and come join us,” Raibert said jokingly to the audience. “Because we’re hiring.”
Filed Under: The Robot Report, Robotics • robotic grippers • end effectors