Good boy! Robotic Dog from Boston Dynamics can now play the world’s slowest fetch game
- The robot company has released a tutorial showing owners how to learn to fetch the four-legged machine
- The retrieval program uses machine learning to help the robot recognize the toy and return it to its owner
- The process requires the owner to create at least 400 images of the toy using the robot to make it work
- The result is a plodding facsimile of the activity favored by the flesh-and-blood inspiration of the machine
Boston Dynamics’ mechanical dog Spot is billed as a utility robot that can perform a variety of tasks from navigating dangerous terrain to carrying heavy loads, and now the robotics company has shown it can also play the world’s least exciting game of fetch.
Using a Spot model equipped with a robotic arm attachment, Boston Dynamics released a tutorial showing how owners of the four-legged machine can teach it to pick up a discarded toy and return it to the thrower.
The process involves using the robot’s machine learning capabilities to recognize and interact with objects in the environment around it, and the tutorial shows users how to train the machine to see the toy.
Robotics company Boston Dynamics demonstrated its robotic dog Spot with retrieving with a rope toy, but at $74,500 per machine, a real dog may be the cheaper, more preferable choice
Spot is billed as a utility machine, capable of carrying loads and navigating challenging terrain
The tutorial asks the user to capture at least 400 images of the rope toy in a specific environment to ensure it can recognize it from different angles.
If they want Spot to play fetch somewhere else, say outside or in the living room, they have to repeat the process for each new environment.
The result is a very slow facsimile of the activity favored by the flesh-and-blood inspiration of the machine, as the robot dog slowly crawls up to the toy, picks it up with its arm, turns and walks over to the owner to finish it. before you retire.
Programming Spot to recognize the toy, as well as the human who throws it, requires the user to capture at least 400 images per environment in which he wants the toy to play
Spot fetch teaching uses the robot’s machine learning capabilities to recognize objects in its environment
The tutorial also shows you how to teach the robot to face the human who threw the toy to return it.
The result is a very slow game of retrieving, but the robot’s creator says it shows the machine’s capabilities in other tasks, such as cleaning up trash.
The company recognizes that someone might have more fun fetching with a living and breathing dog – the Spot has a top speed of 3 miles per hour, and at $74,500 per robot, plus the cost of the robotic arm attachment, a real dog seems like the cheaper and better choice for the game – but notes that learning to fetch the robot demonstrates some potential uses.
“Of course we don’t expect Spot to play fetch in the park on a regular basis, but the same features serve as the foundation for more practical applications, such as automated roadside garbage disposal,” said Andrew Barry, a senior roboticist on the Spot manipulation team.
“You could teach Spot to spot litter (unlike other objects in the area), pick it up, and take it to a trash can. Dog toys are just a starting point for a range of possibilities in the real world.”