Robots vs. animals: Who will win the race in the natural environment? – Neuroscience News

Summary: Scientists have investigated whether modern robots can surpass biological organisms in speed and agility. The study concluded that despite advances in engineering, animals still outperform robots in locomotive efficiency in the natural environment.

The researchers found that the integration of robotic components does not achieve the system-level cohesiveness seen in animals. This insight is fueling a push to develop more adaptive and integrated robotic systems, drawing inspiration from nature’s design.

Key facts:

1. Robotic vs. biological effectiveness: The study confirms that individual robotic subsystems such as performance and control can match or outperform their biological counterparts, but when these systems are combined, robots do not perform as well as animals.
2. Inspiring biological models: Research highlights how animals such as wolf spiders and cockroaches excel in complex terrains and tasks due to their integrated and versatile biological systems.
3. Future engineering directions: The findings will encourage engineers to rethink robot design and advocate a more integrated approach similar to biological systems, where different functions are combined within individual components.

Source: University of Colorado

The question might be a 21st-century version of the fable of the tortoise and the hare: Who would win in a foot race between a robot and an animal?

In a new perspective paper, a team of engineers from the United States and Canada, including University of Colorado Boulder roboticist Kaushik Jayaram, set out to answer this puzzle.

The group analyzed data from dozens of studies and came up with a resounding “no.” In almost all cases, biological organisms such as cheetahs, cockroaches, and even humans seem to be able to outrun their robotic counterparts.

The researchers, led by Samuel Burden of the University of Washington and Maxwell Donelan of Simon Fraser University, published their findings last week in the journal. Scientific robotics.

“As an engineer, it’s a little disconcerting,” said Jayaram, an assistant professor in the Paul M. Rady Department of Mechanical Engineering at CU Boulder. “In 200 years of intensive engineering, we have been able to send spacecraft to the Moon and Mars and much more. But it’s confusing that we don’t yet have robots that are significantly better than biological systems in motion in natural environments.”

He hopes the study will inspire engineers to learn how to build more adaptable and nimble robots. The researchers concluded that the failure of the robots to escape from animals is not due to the lack of a single machine, such as batteries or actuators. Instead, where engineers can falter is making sure these parts work together effectively.

This pursuit is one of Jayaram’s main passions. His lab on the CU Boulder campus is home to plenty of creepy-crawlies, including some hairy wolf spiders about the size of half a dollar.

“Wolf spiders are natural hunters,” Jayaram said. “They live under rocks and can run across difficult terrain at incredible speeds to catch prey.”

He imagines a world in which engineers build robots that function a little more like these extraordinary arachnids.

“Animals are, in a sense, the embodiment of this ultimate principle of design — a system that works really well together,” he said.

Cockroach energy

The question “who can run better, animals or robots?” it’s complicated because running itself is complicated.

In previous research, Jayaram and his colleagues at Harvard University designed a series of robots that try to mimic the behavior of the oft-maligned cockroach. The HAMR-Jr team model fits on a dime and sprints at the equivalent speed of a cheetah. But Jayaram noted that while the HAMR-Jr can rock forward and backward, it doesn’t move as well from side to side or over bumpy terrain.

In contrast, humble cockroaches have no problem running over surfaces from porcelain to clay and gravel. They can also break walls and squeeze through small cracks.

To understand why such versatility remains a challenge for robots, the authors of the new study divided these machines into five subsystems including power, frame, control, sensing and control. To the group’s surprise, few of these subsystems appeared to lack their animal equivalents.

For example, high-quality lithium-ion batteries can deliver up to 10 kilowatts of power for every kilogram (2.2 pounds) they weigh. Animal tissue, on the other hand, produces approximately one-tenth. Muscles, meanwhile, can’t come close to the sheer torque of many engines.

“But at the system level, robots are not that good,” Jayaram said. “We encounter natural compromises in the field of design. If we try to optimize for one thing, like forward speed, we can lose something else, like cornering ability.”

Spider senses

So how can engineers build robots that, like animals, are more than the sum of their parts?

Animals, Jayaram noted, are not divided into separate subsystems in the same way that robots are. For example, your quadriceps power your legs like the HAMR-Jr controllers move your limbs. But the quads also produce their own energy by breaking down fats and sugars and incorporating neurons that can sense pain and pressure.

Jayaram thinks the future of robotics may lie in “functional subunits” that do the same thing: Why not integrate them into a single part rather than separate power sources from your motors and circuit boards?

In a 2015 paper, CU Boulder computer scientist Nikolaus Correll, who was not involved in the current study, proposed such theoretical “robotic materials” that function more like your quad bikes.

Engineers are still far from achieving this goal. Some, like Jayaram, are taking steps in this direction, such as through his lab’s CLARI (Compliant Legged Articulated Robotic Insect), a multi-legged robot that moves a bit like a spider.

Jayaram explained that CLARI relies on a modular design in which each of its legs functions as a separate robot with its own motor, sensors and control circuitry. The team’s new and improved version, called mCLARI, can move in all directions in tight spaces, a first for four-legged robots.

It’s another thing engineers like Jayaram can learn from those consummate hunters, the wolf spiders.

“Nature is indeed a useful teacher.”

About this robotics and neurotech research

Author: Daniel Strain
Source: University of Colorado
Contact: Daniel Strain – University of Colorado
Picture: Image is credited to Neuroscience News

Original Research: Open access.
“Why Animals May Outrun Robots” by Kaushik Jayaram et al. Scientific robotics

Abstract

Why animals can outrun robots

Animals are much better at running than robots. The difference in performance lies in the important dimensions of agility, range and robustness.

To understand the root causes of this performance gap, we compare natural and man-made technologies in five subsystems critical to running: power, frame, control, sensing, and handling.

With few exceptions, technical technologies meet or exceed the performance of their biological counterparts.

We conclude that biology’s advantage over engineering results from better integration of subsystems, and we identify four fundamental hurdles that roboticists must overcome.

Toward this goal, we highlight promising lines of research that have excessive potential to help future running robots achieve animal-level performance.