Researchers at the Massachusetts Institute of Technology have developed a system called RoboGrammar that can test designs Robotic Different by default and select the best based on available parts and local terrain.
There are several ways to find out the most efficient design of a robot for crossing different terrains, so that all different types of robots can simply be built and tested in the real world, but this method is time consuming.
The researchers tell the RoboGrammar system what parts of the robot are around, including things like wheels, joints, etc.
The researchers establish the type of terrain the robot needs to navigate, while RoboGrammar does the rest, so that the computer system generates an optimized structure with the software needed to control the robot.
And theHe said (Alan Zhao) Allen Zhao, lead author of the paper: Robot designs tend to be very similar in shape or design, despite the variety of tasks that robots are used for.
Zhao explains that robot design is a very manual process, and notes that the RoboGrammar project represents a way to develop creative new robotic designs.
Chow says that robotics designers usually turn to four-legged robots when they need to cross different terrains, but researchers questioned whether this design was optimal.
RoboGrammar was designed as a computer model for designing robots because it is not overly influenced by prevailing traditions and perspectives.
The team developed what they call “the rules of graphs”, saying: The result will be a mixture of parts if the computer system links the parts in random ways.
And the rules included things like that parts of the adjacent leg must be connected to one joint and not another part of the leg.
The rules were meant to ensure that computer designs worked, and the team was inspired The rules that control the system of arthropods.
It is reported that the system is not yet ready to allow computers to design robots without any human intervention, but it represents a first step on the road to the possibility of creating environmentally friendly robots.
The team’s next step is to build some robots that the system has molded to see if the simulation’s promises match reality.
“The system can benefit engineers as well as video game designers who need to build populated environments quickly and efficiently,” Zhao said.