CHANNELS

Aerospace

Automotive

Automated & Autonomous Vehicles

Battery & Electrification Technology

Defense

Drone TV

Electronics

Manufacturing & Prototyping

Materials

Medical

Motion Control

RF & Microwave Technology

Robotics & Automation

Sensors & IoT

Test & Measurement

Aerospace
Automotive
Automated & Autonomous Vehicles
Battery & Electrification Technology
Defense
Drone TV
Electronics
Manufacturing & Prototyping
Materials
Medical
Motion Control
RF & Microwave Technology
Robotics & Automation
Sensors & IoT
Test & Measurement
Aerospace & Defense
Search Results

Autonomous Grasping Robots to Help Maintain Space Habitats

Watch this video to see how Purdue University researchers are using machine learning to train a robot to recognize a jumbled pile of items, locate the one item it needs, and retrieve it in the most efficient way.

“It’s easy for robots to manipulate things if the environment is set up perfectly, like at an Amazon warehouse,” said David Cappelleri  , professor of mechanical engineering. “But in a dynamic environment like a human habitat, objects will sometimes be all over the place, or hidden underneath other objects. A robot needs to be smart enough to not only locate an object, but also position itself to successfully grasp and manipulate the object.”

Topics:
Automation Motion Control Robotics

More From SAE Media Group

    Transcript

    00:00:00 So we're trying to study in general the mobile  manipulation problem. So if you have a mobile   robot, it needs to move around the workspace and  pick up different objects to perform different   tasks. Depending on the task and the objects that  are there, you may want to grip it different ways.   So we're trying to understand, you know, which  is the best gripper to use for a particular task,   and then once we know that, what's the best  way to manipulate the gripper to figure out   the perfect orientation for it to grasp the part,  where to grasp it, and then to execute that task.  -We created a virtual environment with all the  objects randomly dropped. We combined different   objects with complex geometries and shapes and  textures. For example we have bananas, apples,   bottles, light bulb, some random cylinders,  or some random 3D printed parts. So what we   want to do then is to create a challenging  environment. Hopefully the robot can learn  

    00:00:53 from this challenging environment and solve  the similar question in the real world.  -This work that we've been doing is for  space. We want to go back to the Moon,   we want to go to Mars. In order to get there,  we need to set up some habitats. And most of   the time these habitats will be unmanned. So the  astronauts will not be there. There will be robots   that will need to be there to set everything up  for them: do maintenance tasks, do repair tasks,   setting up experiments for them, clearing cargo,  putting cargo away. But these robots need to be   able to identify the objects that they need to  manipulate, figure out the best way to manipulate   them to accomplish the task. When the humans  are there, the astronauts are there, we want   these robots to actually work with the humans  as well, to jointly carry out these different   kind of tasks. And so this could be a nice  assistant for the astronauts while they're there.