Look out, athletes — robots can now play sports. These machines use an assortment of legs, algorithms or actuators to make complex movement possible. While this field is still in its infancy, engineers have already made numerous advancements to existing technology.
There are numerous benefits to designing and deploying sports robots. In addition to furthering robotics innovations, it drives curiosity in science, technology, engineering and math (STEM). If more people — especially young individuals — see these novel creations and their interests converge, they may be more inclined to enter the field.
Improved training is another significant benefit of integrating robotics into sports. One research team discovered the human brain responds to machine opponents very differently when analyzing how one person’s neurons acted while they played against a table tennis robot. Coaches could use this technology to push their athletes to greater heights.
Another benefit of sports robots is marketability. Since they can perform tasks faster and more consistently than humans, they’re ideal for entertainment. Unlike high-performing athletes, they never get tired, don’t require salaries and can’t get injured. Event coordinators could use them for sporting contests, competitions and games.
Sports robots can only further robotics advancement, drive interest in STEM or become a staple in stadiums if they operate as intended. If a hacker infiltrates a machine’s connected systems or networks, they could cause behavioral faults or even permanent damage.
Sporting events are an ideal target for threat actors seeking to gain publicity, steal money, or stir terror because of their media coverage and profitability. Considering cyber attacks are increasing in frequency and severity, further expanding the attack surface with robotics integrations without first addressing security risks would be unwise.
Mechanical engineers should inventory every component, network, data storage system, and software their robot uses to assess risk and determine their vulnerability to various cyber threats. Once they understand the likelihood of cyber threats, they should figure out what their machine-specific indicators of compromise look like — whether that’s a slow connection, abnormal network logs or motors turning on when they shouldn’t. This knowledge can help them determine when a bad actor has infiltrated their system.
Event coordinators should work with engineering teams to keep robots off public networks when participating in sporting events. They should be segmented away from other devices to prevent lateral movement.
There are numerous sports-playing robots in existence.
Researchers from Georgia Tech developed a tennis-playing robot after deciding an upgrade for the stationary ball feeder is long overdue. They named their creation ESTHER in homage to wheelchair tennis player Esther Vergeer.
ESTHER comprises a fully autonomous, multi-directional arm, a wheelchair base and a tennis racket. In addition to its built-in computer vision, it relies on strategically placed cameras around the court to triangulate the ball’s trajectory.
The algorithm trained extensively on sample throws to calculate where to move and how to return the ball. Now that the researchers have successfully demonstrated its tennis-playing capabilities, they’re working on teaching its AI to strategize and win matches.
Cassie is a bipedal robot from the Oregon State University College of Engineering designed to run long distances. When the engineers began tweaking its stride, they wondered how fast it could go. They pursued that question — and their creation eventually set the world record for the fastest 100-meter sprint by a bipedal robot at 24.73 seconds.
The engineering team trained Cassie for the equivalent of one year in only one week using simultaneous training experiences — a technique called parallelization. Their creation has no external cameras or sensors, instead relying on advanced AI.
One of the neural networks was specialized for speed, while the other was custom-made for being stationary. Cassie transitions between the two to move from a standstill to a sprint and vice versa. Without this novel AI approach, it couldn’t quickly achieve high speeds, or stop sprinting and remain stable.
Mechanical engineers from the University of California Los Angeles developed a soccer-playing robot called Advanced Robotic Technology for Enhanced Mobility and Improved Stability (ARTEMIS). It stands at 4 feet, 8 inches tall and weighs 85 pounds.
ARTEMIS can maintain its balance when shoved, withstand the impact of thrown objects, run and evade others and kick a soccer ball. Electric actuators — which are becoming increasingly common because they’re easily programmable and make very little noise — make these feats possible. The engineers modeled them after biological muscle functioning.
While Boston Dynamics — a well-known engineering and robotics firm — first unveiled Atlas in 2013, its fully electric model debuted in 2024. The human range of motion doesn’t constrain this latest iteration, so its gymnastic feats are particularly impressive.
Atlas can flip, vault and somersault. It’s even capable of complex actions like handstands, 360˚ spin jumps and split leaps. Swiveling joints are the functional developments behind these movements.
The firm also developed a unique software combination to enable real-time decision-making. Once an optimization algorithm transforms detailed maneuver descriptions into dynamic reference motions, Atlas uses predictive modeling to go from one motion to the next.
Disney’s Imagineers — the research and development team behind the company’s creative engineering — demoed one of their latest creations at the South by Southwest media festival in 2023. The child-sized robot climbs out of a box, skates across the stage, does a forward roll that imitates a fall and then stands up independently.
Motion capture is the unique technology behind this prototype's fluid, lifelike movements. The Imagineers wanted to integrate a human element into the mechanical performance. At the same time, they tried to embrace failure, meaning the robot was designed to recover from accidents rather than remain stable — an interesting approach for a fall-prone sport.
University of Michigan researchers developed a tri-pedal skating robot. It’s more effective and stable than similar robots because it combines a ball-bot's maneuverability with a legged machine's stability. This novel locomotion form leverages simultaneous pushing and rolling actions.
SKOOTR’s base is a sphere — a hub that houses all its electronic components. Its three robotic legs have two joints and one rubber foot to maximize reach and maneuverability. The servo switches between gripping or passive rolling, granting it more stability than similar tri-pedal robots.
Human athletes are safe for now — most sports robots are proofs of concept to help progress robotics applications in areas like defense, search and rescue, or entertainment. However, there’s a possibility these machines may one day supplement athletes as technology progresses.