Harvard and Adelaide researchers develop bio-inspired swarm robots

Harvard and Adelaide researchers develop bio-inspired swarm robots

Researchers from Harvard University and Adelaide University have developed decentralized robotic swarms that mimic the collective intelligence of social insects to perform complex tasks without a central controller. These systems, inspired by the behaviors of ants and bees, are designed to operate autonomously in harsh or unpredictable environments, such as planetary exploration missions.

Traditional automated systems often rely on a single control center, which can create a point of failure that renders an entire operation inflexible or vulnerable. In contrast, these bio-inspired swarms make independent decisions while working collaboratively. This ensures the system continues to function even if individual robotic units fail.

The Harvard "RAnts" Model

At the Harvard John A. Paulson School of Engineering and Applied Sciences (SEAS) and the Faculty of Arts and Sciences, a team led by Professor L. Mahadevan created a fleet of cooperative robots called RAnts. These robots use a process called stigmergy, where individuals modify their environment and then respond to those modifications.

While ants use chemical pheromones to coordinate, RAnts utilize photormones — light fields that serve as digital substitutes for chemicals. The robots follow these light gradients, transporting building blocks and depositing material once specific signal thresholds are met.

"Our new study shows how simple, local rules can lead to the emergence of complex task completion that is self-organized and thus robust and adaptive,"

L. Mahadevan, Professor at SEAS and FAS, via eurekalert.org

Mahadevan introduced the concept of exbodied intelligence, suggesting that collective cognition comes from the interaction between agents and an evolving environment rather than just the individual robots. The swarm's behavior is tuned via two parameters: deposition rate and cooperation strength. By adjusting these, the robots can switch between building new structures or dismantling existing ones.

According to the study published in PRX Life, these robots form nucleation sites through a mechanism called trapping instability. As robots become temporarily confined by their own signals, others converge on the location, accelerating the construction of organized aggregates.

Adelaide University's Mining Application

Parallel research at Adelaide University focused specifically on the mining industry, where operations are moving into deeper and more remote locations. Using Zumo 2040 robots in a laboratory environment designed to mimic a mine, the team tested different insect-inspired strategies.

The research, published in Natural Sciences, found that a honeybee-inspired approach performed best in all tests. This method involved robots exploring an area and remembering resource locations. An ant-inspired approach also showed improvement by dividing labor, with some robots finding resources while others transported them.

"By applying these ideas to robotics, we can create systems that are more efficient, adaptable, and reliable for industries such as mining,"

Dr. Joven Tan, lead author and PhD researcher at the School of Chemical Engineering, via sflorg.com

Project leader Dr. Noune Melkoumian stated that these systems could be deployed in dangerous or difficult-to-reach mining areas to reduce human risk and increase productivity. She noted that such autonomy will be essential for future space mining missions.

Challenges and Future Implementation

Despite the successful laboratory tests, researchers acknowledge several hurdles before these swarms can be widely deployed in industrial mines. Key challenges include:

• Extending battery life for long-term autonomy.
• Improving sensor capabilities for better navigation.
• Adapting systems to unpredictable underground conditions.
• Developing robust communication systems to ensure reliability.

Ongoing efforts to enhance these systems include the integration of machine learning for autonomous decision-making and the exploration of new materials to ensure durability in harsh environments. The goal is to create an energy-efficient alternative to rigid centralized systems that minimizes the carbon footprint of resource extraction.

The Harvard study was co-authored by S. Ganga Prasath and Fabio Giardina, with funding provided by the Simons Foundation, the Henri Seydoux Fund, and several National Science Foundation grants.