Digit, the bipedal robot of Agility Robotics, has been given a makeover before it applies for jobs in 2024.

Agility Robotics, one of the world’s leading companies in robotics, has long been working on Digit — a bipedal robot made to help in routine warehouse tasks and jobs. The robot is designed to work in warehouse and logistics operations by moving bins from one shelf to a conveyor belt.

While this is a task that is pretty straightforward and Digit can do successfully, Agility Robotics took it upon themselves to upgrade the bipedal robot further by giving it a “makeover”.

What once was a bipedal robot that only had arms, legs, and a torso now has a head for added human-robot interaction (HRI). Although the new design of Digit V4 now includes a horizontally-elongated head, this robot is still considered a bipedal rather than a humanoid. Aside from this, Digit V4 also gets a new set of hands that can help it move bins and totes more efficiently and securely.

Both the design and function of Digit V4 may not seem a lot, but Agility Robotics makes a good point regarding the potential of this bipedal robot.

“Digit is multipurpose, so it can execute a variety of tasks and adapt to many different workflows; a fleet of Digits will be able to switch between applications depending on current warehouse needs and seasonal shifts. Because Digit is also human-centric, meaning it is the size and shape of a human and is built to work in spaces designed for people, it is easy to deploy into existing warehouse operations and as-built infrastructure without costly retrofitting.”.

At the moment of writing, Digit V4 is still limited to the aforementioned design and functions but it seems like Agility Robotics intends to further enhance and upgrade this bipedal robot in time for a mass shipment to its partners for “official employment” in early 2024.

With a run time of 53 minutes and 3 seconds, Cassie the bipedal robot makes history as the first robot to run and complete a 5k course on a single charge.

Although it might be long compared to any human world record, having a robot successfully run a 5k course is impressive on its own considering that Cassie is only a pair of mechanical legs powered by a computer. As of writing, Cassie is also the fastest bipedal robot to run 100 meters with a Guinness World Record to prove it.

Using a deep reinforcement learning algorithm, Cassie was able to figure out how to stay upright and balance herself while running the 5k course in the fields of Oregon State University.

“Deep reinforcement learning is a powerful method in AI that opens up skills like running, skipping, and walking up and down stairs,” says undergrad student Yesh Godse who is part of the team that built Cassie.

To complete the 5k run, Cassie underwent a year’s worth of simulated environment training compressed into a week through a computing technique known as parallelization. This technique enabled Cassie to undergo a series of training experiences simultaneously thus, deep reinforcement learning.

Professor Jonathan Hurst, who is also the Chief Technology Officer at Agility Robotics, lead the team who built and invented Cassie. He marks Cassie’s history and Guinness Record accomplishment as the first of many robotics feats.

“This may be the first bipedal robot to learn to run, but it won’t be the last. I believe control approaches like this are going to be a huge part of robotics. The exciting part of this race is the potential. Using learned policies for robot control is a very new field, and this 100-meter dash is showing better performance than other control methods. I think progress is going to accelerate from here.”, Hurst says.

Cassie was invented and built by a team from the Dynamics Robotics Laboratory in Oregon State University’s College of Engineering after receiving a grant of $1 million from the Advanced Research Projects Agency of the U.S. Department of Defense.

According to the team, Cassie and robots alike will be used for commercial purposes that can help industries in deliveries, warehousing, and even home assistance.