a system called RoboGrammar. The system creates arthropod-inspired
robots for traversing a variety of terrains. Pictured are several
robot designs generated with RoboGrammar.
Credits:Courtesy of the researchers
So, you need a robot that climbs stairs. What shape should that
robot be? Should it have two legs, like a person? Or six, like an
ant?
Choosing the right shape will be vital for your robotâs
ability to traverse a particular terrain. And itâs impossible to
build and test every potential form. But now an MIT-developed
system makes it possible to simulate them and determine which
design works best.
You start by telling the system, called RoboGrammar, which robot
parts are lying around your shop â wheels, joints, etc. You also
tell it what terrain your robot will need to navigate. And
RoboGrammar does the rest, generating an optimized structure and
control program for your robot.
The advance could inject a dose of computer-aided creativity
into the field. âRobot design is still a very manual process,â
says Allan Zhao, the paperâs lead author and a PhD student in the
MIT Computer Science and Artificial Intelligence Laboratory
(CSAIL). He describes RoboGrammar as âa way to come up with new,
more inventive robot designs that could potentially be more
effective.â
Zhao is the lead author of the paper, which he will present at
this monthâs SIGGRAPH Asia conference. Co-authors include PhD
student Jie Xu, postdoc Mina KonakoviÄ-LukoviÄ, postdoc Josephine
Hughes, PhD student Andrew Spielberg, and professors Daniela Rus
and Wojciech Matusik, all of MIT.
Ground rules
Robots are built for a near-endless variety of tasks, yet
âthey all tend to be very similar in their overall shape and
design,â says Zhao. For example, âwhen you think of building a
robot that needs to cross various terrains, you immediately jump to
a quadruped,â he adds, referring to a four-legged animal like a
dog. âWe were wondering if thatâs really the optimal
design.â
Zhaoâs team speculated that more innovative design could
improve functionality. So they built a computer model for the task
â a system that wasnât unduly influenced by prior convention.
And while inventiveness was the goal, Zhao did have to set some
ground rules.
The universe of possible robot forms is âprimarily composed of
nonsensical designs,â Zhao writes in the paper. âIf you can
just connect the parts in arbitrary ways, you end up with a
jumble,â he says. To avoid that, his team developed a âgraph
grammarâ â a set of constraints on the arrangement of a
robotâs components. For example, adjoining leg segments should be
connected with a joint, not with another leg segment. Such rules
ensure each computer-generated design works, at least at a
rudimentary level.
Zhao says the rules of his graph grammar were inspired not by
other robots but by animals â arthropods in particular. These
invertebrates include insects, spiders, and lobsters. As a group,
arthropods are an evolutionary success story, accounting for more
than 80 percent of known animal species. âTheyâre characterized
by having a central body with a variable number of segments. Some
segments may have legs attached,â says Zhao. âAnd we noticed
that thatâs enough to describe not only arthropods but more
familiar forms as well,â including quadrupeds. Zhao adopted the
arthropod-inspired rules thanks in part to this flexibility, though
he did add some mechanical flourishes. For example, he allowed the
computer to conjure wheels instead of legs.
A phalanx of robots
Using Zhaoâs graph grammar, RoboGrammar operates in three
sequential steps: defining the problem, drawing up possible robotic
solutions, then selecting the optimal ones. Problem definition
largely falls to the human user, who inputs the set of available
robotic components, like motors, legs, and connecting segments.
âThatâs key to making sure the final robots can actually be
built in the real world,â says Zhao. The user also specifies the
variety of terrain to be traversed, which can include combinations
of elements like steps, flat areas, or slippery surfaces.
With these inputs, RoboGrammar then uses the rules of the graph
grammar to design hundreds of thousands of potential robot
structures. Some look vaguely like a racecar. Others look like a
spider, or a person doing a push-up. âIt was pretty inspiring for
us to see the variety of designs,â says Zhao. âIt definitely
shows the expressiveness of the grammar.â But while the grammar
can crank out quantity, its designs arenât always of optimal
quality.
Choosing the best robot design requires controlling each
robotâs movements and evaluating its function. âUp until now,
these robots are just structures,â says Zhao. The controller is
the set of instructions that brings those structures to life,
governing the movement sequence of the robotâs various motors.
The team developed a controller for each robot with an algorithm
called Model Predictive Control, which prioritizes rapid forward
movement.
âThe shape and the controller of the robot are deeply
intertwined,â says Zhao, âwhich is why we have to optimize a
controller for every given robot individually.â Once each
simulated robot is free to move about, the researchers seek
high-performing robots with a âgraph heuristic search.â This
neural network algorithm iteratively samples and evaluates sets of
robots, and it learns which designs tend to work better for a given
task. âThe heuristic function improves over time,â says Zhao,
âand the search converges to the optimal robot.â
This all happens before the human designer ever picks up a
screw.
âThis work is a crowning achievement in the a 25-year quest to
automatically design the morphology and control of robots,â says
Hod Lipson, a mechanical engineer and computer scientist at
Columbia University, who was not involved in the project. âThe
idea of using shape-grammars has been around for a while, but
nowhere has this idea been executed as beautifully as in this work.
Once we can get machines to design, make and program robots
automatically, all bets are off.â
Zhao intends the system as a spark for human creativity. He
describes RoboGrammar as a âtool for robot designers to expand
the space of robot structures they draw upon.â To show its
feasibility, his team plans to build and test some of
RoboGrammarâs optimal robots in the real world. Zhao adds that
the system could be adapted to pursue robotic goals beyond terrain
traversing. And he says RoboGrammar could help populate virtual
worlds. âLetâs say in a video game you wanted to generate lots
of kinds of robots, without an artist having to create each one,â
says Zhao. âRoboGrammar would work for that almost
immediately.â
One surprising outcome of the project? âMost designs did end
up being four-legged in the end,â says Zhao. Perhaps manual robot
designers were right to gravitate toward quadrupeds all along.
âMaybe there really is something to it.â
Originally published by
Daniel Ackerman | MIT News
Office | November 30, 2020
MIT