• Dr. Timothy Smith

Swarm Robotics


Photo Source: Wikimedia Commons

Many robots have been designed to look and act like people because we often are looking for ways to automate what people do, and it is a form which people easily relate to. However, the field of robotics and artificial intelligence also gets inspiration from other animals in the natural world. Although many people find insects creepy and even frightening when thinking of them in swarms such as fire ants or killer bees, entomologists study insects and continue to seek to understand how ants and bees function as a collective group with no central organizing leader. In most cases of social insects, the colony serves the queen, but the queen does not have a group of dedicated administrators that tell each ant or bee what to do. In the case of ants, individuals will respond to their environment and to chemical signals called pheromones emitted by other ants they encounter. For example, an ant that has returned from a food source will emit a pheromone combination that tells another ant to start gathering food. The second ant without food will follow a trail of scent back to the food source laid down by the first ant. Ants are versatile, and each ant can perform tasks like foraging, shuttling baby ants around, nest building, and more. Following simple signs from one individual to another gives the colony a unique ability to thrive without instructions or commands coming from some command center. The term swarm intelligence describes the ability of ants to build massive colonies, avoid danger and invade whole new continents without a central leader.

Swarm intelligence refers to the amplified intelligence of groups of animals. Birds form flocks, fish form shoals, and ants form colonies. An animal like a fish or an ant may not be brilliant individually, but animal swarms avoid predators or build magnificent homes like the termite colony or honey bee hive. The nature of swarm intelligence not only intrigues biologists but has captured the imagination of computer scientists and robotics engineers as well. Inspired by the collective intelligence of a colony of ants or bees, research continues into developing simple robots that work together collectively. The idea that simple robots that react to their environment and each other as a collective drives research today and forms a new branch of artificial intelligence. Swarm sounds frightening and menacing, but some applications being developed today that use swarm intelligence may prove very helpful in areas such as agriculture and healthcare. In the agriculture industry, David Dorhout believes that the future of farming may depend on armies of cheap little robots equipped with sensors and some basic rules to guide them to replace massive farm machinery. He envisions little robots using swarm intelligence to plant, tend, and harvest various crops. There will not be some complicated central planning program that tells the robots were to go and what to plant. Instead, the robots will sense where the other robots are and move to open spaces to plant seeds and detect when moisture is low and water individual plants. Mr. Dorhout calls his company Dorhout R&D and his farm robot Prospero. The medical community also envisions employing swarm intelligence at the level of nanoparticles for treating cancer. The tiny nanoparticles pass through a blood vessel into cancerous tissue and recognize a tumor and attack it. Swarm intelligence may also contribute to making autonomous cars safer by letting cars talk to each other locally to prevent collisions and smooth traffic flow.

Clearly on the more menacing side, military applications of swarm robotics may come in many forms. The military frequently leads the way in technological advancement and already supports numerous projects developing robotics and swarm technology. According to an article in New Scientist, the US military envisions using robot swarms in removing mines and for search and rescue without human intervention. Recently the US Navy released a video on YouTube of F/A 18 Hornet fighter jets launching just over 100 mini-drones over the desert and the subsequent testing of the drone swarm (posted below). Unlike typical drones that have one operator per drone, the drone swarm uses a collective brain shared among all the drones that communicate and work like a swarm in nature. The drones are battery powered, 12-inch-long aircraft named Perdix and were developed by students at Massachusetts Institute of Technology’s Lincoln Labs made of commercially available and 3-D printable parts. The video also shows the swarm clustering and then moving to patrol an area. They look and sound like a menacing swarm of huge bees or hornets. It seems almost ironic that the Hornet was used to launch in flight the drones from underwing pods. The Perdix drones may be initially helpful for surveillance in battlefield situations. By extension, their small size, the large number, and low cost could also make them a new and challenging to defend against weapons.

Swarm intelligence observed in nature in the form of ant and bee colonies demonstrate how the actions of individuals together without a central organizer amplify their collective intelligence that is smarter than any individual member of the colony. The collective intelligence observed in the natural world inspired a new discipline in artificial intelligence that seeks to use cooperative behavior to empower simpler robots to work as a team for a common goal.

Dr. Smith’s career in scientific and information research spans the areas of bioinformatics, artificial intelligence, toxicology, and chemistry. He has published a number of peer-reviewed scientific papers. He has worked over the past seventeen years developing advanced analytics, machine learning, and knowledge management tools to enable research and support high level decision making. Tim completed his Ph.D. in Toxicology at Cornell University and a Bachelor of Science in chemistry from the University of Washington.

You can buy his book on Amazon in paperback here and in kindle format here.