Close

Stanford's New Spiny Grippers Will Help RoboSimian Go Rock Climbing

A new microspine gripper can hold enough weight to allow JPL's quadruped to go rock climbing

3 min read
Microspine gripper for robots
Photo: Stanford

Over a decade ago, Stanford roboticists started experimenting with ways of using arrays of very small spines to help climbing robots grip rough surfaces. These microspine grippers have been used on all kinds of research robots since then, and recently, NASA has decided that microspines are the best way for spacecraft to grab onto asteroids.

Yesterday at the IEEE/RSJ International Conference on Intelligent Robots and Systems in South Korea, Shiquan Wang from Stanford presented a new microspine-based palm design for rock-climbing robots. These palms use microspines that can support four times the weight of previous designs, which will be enough to turn JPL’s RoboSimian DRC robot into a champion rock climber. And we’re not talking just scrambling up slopes: It’ll be able to scale vertical rock faces, and even clamber around overhangs. 

Microspines work just like tiny little claws. They catch and hold onto rough surfaces, and while each spine is itty bitty and can’t hold much, if you use enough of them, you can support a lot of weight (or resist a lot of force). The previous generation of microspines, which include the ones that NASA is using for its Asteroid Redirect Mission, are highly compliant, able to hook onto very rough surfaces by allowing each spine to find its own little spot to grab onto. The compliant design is very robust for lots of applications, but the compliant mechanism is bulky, and reduces the amount of spines you can stuff into your gripper.

Microspines

Photo: Evan Ackerman/IEEE Spectrum
Close-up view of the microspine gripper.

If your goal is to support as much weight as possible for applications like rock climbing, you want to have as many spines sticking into the surface and supporting weight as you can, and this is where Stanford’s new gripper design comes from: by removing almost all of the compliance, you can vastly increase your spine density. The overall percentage of spines that are engaged onto the surface goes down, but since you have so many more spines, you still end up way, way ahead on how much load you can handle.

Robot microspines gripper

Image: Stanford
The illustrations on the left show a comparison between the earlier spine mechanism design (top) and the new, linearly-constrained design (bottom). The photos on the right show a spine tile (top) and an individual spine with its miniature spring (bottom).

The design of these new spines is straightforward: each 15-mm steel spine sits in a 3D printed sleeve, with a spring that pushes it down towards the surface it’s trying to grip. That single compliant axis helps the spines grip very rough surfaces, and 60 spines will fit into a single 18-mm x 18-mm “tile.” Twelve tiles together form this prototype palm, and each tile has a little bit of wiggle room to help it engage better to optimize load sharing. All of the spines are angled slightly towards the surface that the palm is gripping, meaning that they engage when a force is pulling on the palm, but lifting in the opposite direction causes the palm to easily detach.

The researchers tested the full palm on nine different surfaces and achieved up to 710 N of shear adhesion, which is over four times better than previous designs. It doesn’t like surfaces that are super smooth or super rough, but for most rock surfaces (including concrete), it works great. Next, the palm design will expand to include compliant fingers and toes, each covered with microspine tiles, and they’ll see what JPL’s RoboSimian can do with them. Look for that to happen next year.

“A Palm for a Rock Climbing Robot Based on Dense Arrays of Micro-Spines,” by Shiquan Wang, Hao Jiang, and Mark R. Cutkosky from Stanford University, was presented yesterday at the 2016 IEEE/RSJ International Conference on Intelligent Robots and Systems in Daejeon, Korea.

The Conversation (0)

How the U.S. Army Is Turning Robots Into Team Players

Engineers battle the limits of deep learning for battlefield bots

11 min read
Robot with threads near a fallen branch

RoMan, the Army Research Laboratory's robotic manipulator, considers the best way to grasp and move a tree branch at the Adelphi Laboratory Center, in Maryland.

Evan Ackerman
LightGreen

This article is part of our special report on AI, “The Great AI Reckoning.

"I should probably not be standing this close," I think to myself, as the robot slowly approaches a large tree branch on the floor in front of me. It's not the size of the branch that makes me nervous—it's that the robot is operating autonomously, and that while I know what it's supposed to do, I'm not entirely sure what it will do. If everything works the way the roboticists at the U.S. Army Research Laboratory (ARL) in Adelphi, Md., expect, the robot will identify the branch, grasp it, and drag it out of the way. These folks know what they're doing, but I've spent enough time around robots that I take a small step backwards anyway.

The robot, named RoMan, for Robotic Manipulator, is about the size of a large lawn mower, with a tracked base that helps it handle most kinds of terrain. At the front, it has a squat torso equipped with cameras and depth sensors, as well as a pair of arms that were harvested from a prototype disaster-response robot originally developed at NASA's Jet Propulsion Laboratory for a DARPA robotics competition. RoMan's job today is roadway clearing, a multistep task that ARL wants the robot to complete as autonomously as possible. Instead of instructing the robot to grasp specific objects in specific ways and move them to specific places, the operators tell RoMan to "go clear a path." It's then up to the robot to make all the decisions necessary to achieve that objective.

Keep Reading ↓ Show less