Energy-Efficient ‘SlothBot’ Moves Only When It Has To

Billy Hurley, Digital Editorial Manager

While robots like the WildCat from Boston Dynamics reach speeds of just under 20 miles an hour, engineers from Georgia Tech have gone with a decidedly slower approach.

Taking inspiration from a mammal that moves at a leisurely pace, the researchers created a proof-of-concept SlothBot.

“In robotics, it seems we are always pushing for faster, more agile and more extreme robots,” said Magnus Egerstedt , the Steve W. Chaddick School Chair of the School of Electrical and Computer Engineering at the Georgia Institute of Technology and principal investigator for SlothBot. “But there are many applications where there is no need to be fast. You just have to be out there persistently over long periods of time, observing what’s going on.”

Operating only when necessary, the energy-efficient robot supports the long-term surveillance tasks required in precision agriculture, infrastructure maintenance, security, and environmental monitoring.

Powered by a pair of photovoltaic panels, the SlothBot is designed to linger in forest canopy continuously for months. SlothBot moves only to measure environmental changes, such as weather and chemical factors.

Mechanically, SlothBot consists of two bodies connected by an actuated hinge. Each body houses a driving motor connected to a tire.

The proof-of-concept robot, described May 21 at the International Conference on Robotics and Automation (ICRA) in Montreal , operates on a network of treetop cables.

Graduate Research Assistant Gennaro Notomista designed SlothBot together with his colleague, Yousef Emam, using 3D-printed parts for the gearing and wire-switching mechanisms needed to crawl through the tree wires — and switch from one cable to another without falling.

“The challenge is smoothly holding onto one wire while grabbing another,” said Notomista. “Making sure the switches work well over long periods of time is really the biggest challenge.”

The Slothbot is currently only crawling around cables on the Georgia Tech campus.

The researchers, however, hope to soon see the robot monitoring conditions in the Atlanta Botanical Garden and even a cacao plantation in Costa Rica.

"The cables used to move cacao have become a sloth superhighway because the animals find them useful to move around,” Egerstedt said. “If all goes well, we will deploy SlothBots along the cables to monitor the sloths.

With technology, it’s not always about achieving the highest speed. Egerstedt spoke with Tech Briefs about the importance of embracing slowness.

Tech Briefs: When people think of technology imitating nature, they may overlook the benefits of a robot being sloth-like. What are the benefits of a robot that acts like a sloth?

Magnus Egerstedt: Sloths, like almost all arboreal folivores (tree-dwelling, leaf-eating animals), live highly energy-constrained lives. As a result, they have to be extremely strategic about how they spend their energy. This also means that they only move when they absolutely have to.

In my research, I am interested in so-called long-duration autonomy problems where robots have to perform over very long time scales and where “survival” — don’t run into obstacles, don’t get stranded somewhere without any chance of recharging the batteries, don’t get stuck — matters more than whatever primary objective the robot is supposed to address. In environmental monitoring applications, such as robots on farm fields inspecting the health of crops, nothing has to happen at high speeds, and being energy efficient is more important than being fast and agile. Hence the sloth…

Tech Briefs: What inspired you and your team to use sloth-inspired design? Did it feel counter-intuitive to go with a slower animal?

Egerstedt: This work started when I visited Costa Rica and got very interested in sloths. In particular, I couldn’t understand how they could survive and not be instantly eaten by predators. So, I started reading up on sloths and even got in touch with a sloth ecologist, Jon Pauli at the University of Wisconsin. What I learned is that in ecology, richness of behavior is driven by constraints as opposed to goals. Sloths move slowly because constraints on the (lack of) availability of food forces this behavior.

And this idea of letting constraints, as opposed to goals, drive the robot behaviors is ultimately what led to the SlothBot. And yes, it is very counterintuitive to embrace slowness as a design paradigm. I’ve gotten quite a few skeptical comments from fellow roboticists.

Tech Briefs: What kinds of criticism have you heard?

Egerstedt: I wouldn’t exactly call it criticism. It is more that people are a bit surprised to see slowness being embraced explicitly since it goes against what we, as a community, are aiming for in our research. My response typically falls back on the wonderful heterogeneity we find in nature. There are fast animals, and there are slow animals. And they clearly fit a particular ecological niche. It would be rather strange if robots shouldn’t likewise be different in response to different tasks and environments.

Tech Briefs: How exactly does the SlothBot move from one cable to the next? That seems like a tricky maneuver.

It is tricky. We came up with a design where we have a gear wheel that can rotate with an opening in it. And by rotating, the different cables get moved into or out of the wheel as a way of grabbing and releasing the cables. (see figure below)

Tech Briefs: What parts of the technology enable the robot to look more like a sloth?

Egerstedt: The “slothy” aspects of the design are mainly behavioral as opposed to esthetic, i.e., the robots act like sloths more than look like them. But, as we are about the deploy the SlothBot in the Atlanta Botanical Garden, we realized that we need to make it look more like a sloth as well. We have now developed a body that makes the robot look like a high-tech sloth. But, in all honesty, this is more for show than anything else.

Tech Briefs: Where have you tested the SlothBot and how has it performed?

Egerstedt: The Slothbot has only been tested on Georgia Tech’s campus. So far, the SlothBots have performed as expected and have never run out of battery while being unable to move to a sunny spot to recharge.

Tech Briefs: What are its limitations?

Egerstedt: During the design phase, we decided to make the robot cable-driven since a flying robot requires too much energy and a ground robot can easily get stuck. The main limitation is that the robot requires a cable-network infrastructure for it to be able to operate.

Tech Briefs: What applications are possible because of the technology’s sloth-like nature?

Egerstedt: I am very interested in environmental monitoring applications, broadly speaking. A few years ago I read about how someone claimed to have spotted the presumed extinct Ivory-Billed Woodpecker in the dense forests of Louisiana. I would love to be able to deploy a team of SlothBots to see if they can find this fabled bird. In fact, there are a lot of questions pertaining to environmental conservation that we do not have the answers to, like what animals are pollinating certain orchids. The hope is that the SlothBot will be able to play a key role in uncovering some of these answers, which are increasingly important given the changing climate.

What do you think of the SlothBot? Share your comments and questions below.

---