Google DeepMind researchers have just developed a standardized test to assess the performance of four-legged robots. Inspired by dog competitions, they designed an obstacle course called… Precure.
You will also be interested
[EN VIDÉO] When the Spot robot reproduces the choreography… The Boston Dynamics Spot robot reproduces the choreography for the “Start Me Up…
Robots have evolved a lot in recent years, gaining speed and agility. One of the most popular forms is the four-legged dog or robot, like Boston Dynamics says. However, how do you compare different models when there is no standard to measure their performance?
To solve this problem, researchers from Google DeepMind I decided to create an obstacle course intended to serve as a benchmark to measure the robots’ performance. Baptized parkour, contraction barking (barking in English) f parkourinspired by agility courses for dogs.
Robots are half as fast as dogs
The court is part of a square of 5 meters on each side, containing start and finish tables, a slalom, a barrier (or A-frame) with ups and downs, and a long jump. They initially tested precure with “dooglers,” Google employees’ dogs, who were able to complete the course in 10 seconds. The robotic dogs used in the test took twice as long.
The researchers then score the performance between setting an agility score between 0 and 1. It is calculated by taking the maximum score, which is 1, and subtracting a penalty of 0.1 for each obstacle failure, and also for the time spent over 10 seconds divided by 100, which is 0.1 for a robot that takes 20 seconds in total.
The researchers hope that Parkour can be used as a benchmark test to compare different robot models, or to evaluate the performance of the same robot with different sensors, in order to eventually achieve animal-level agility.
