Cassie the bipedal robot uses machine learning to complete a 5km jog
Four years is a long time in robotics, especially so for a bipedal robot developed at Oregon State University (OSU) named Cassie. Dreamt up as an agile machine to carry packages from delivery vans to doorsteps, Cassie has recently developed an ability to run, something its developers have now shown off by having it complete what they say is the first 5-km (3.1-mi) jog by a bipedal robot.
We first took a look at Cassie the bipedal robot back in 2017, when OSU researchers revealed an ostrich-like machine capable of waddling along at a steady pace. It is based on the team's previously developed Atrias bipedal robot, but featured steering feet and sealed electronics in order to function in the rain and snow and navigate outdoor terrain.
The team has since used machine learning to equip Cassie with an impressive new skill: the ability to run. This involved what they call a deep reinforcement learning algorithm, which Cassie combines with its unique biomechanics and knees that bend like an ostrich to make fine adjustments to keep itself upright when on the move.
“Deep reinforcement learning is a powerful method in AI that opens up skills like running, skipping and walking up and down stairs,” says team member Yesh Godse.
Running robots are of course nothing new. Honda's ASIMO robot has been jogging along at speeds of up to 6 km/h (3.7 mph) since 2004, and in 2011 we looked at machine called Mabel with a peak pace of 10.9 km/h (6.8 mph), which was billed as the world's fastest bipedal robot with knees. More recently, the Atlas humanoid robot from Boston Dynamics has wowed us not just by running through the woods, but performing backflips and parkour.
The OSU team were keen to show off the endurance capabilities of Cassie, by having it use its machine learning algorithms to maintain balance across a 5-km run around the university campus, while untethered and on a single charge of its batteries. It wasn't all smooth sailing, with Cassie falling down twice due to an overheated computer and a high-speed turn gone wrong. But following a couple of resets, the run was completed in a total time of 53 minutes and 3 seconds.
“Cassie is a very efficient robot because of how it has been designed and built, and we were really able to reach the limits of the hardware and show what it can do,” said Jeremy Dao, a Ph.D. student in the Dynamic Robotics Laboratory.
According to the researchers, this is the first time a bipedal robot has finished a 5-km run, albeit it took place at walking speed and needed a little help along the way. It is possible that other bipedal robots may be capable of covering such distances, but it is also possible that no one has thought to try. Either way, the run is an impressive demonstration of the progress being made by the team. Check it out below.
Source: Oregon State University