Manas Verma elucidates upon the concept of ‘Robot Swarms’ and comments upon its feasibility and benefits. He also explains the concept with the help of day to day examples from animal as termites, equating their behavior with the possible conduct of the robot swarms, if successfully created and implemented.
He puts an account of the significance of robot swarms in the modern scenario.
Arguably one of the most futuristic styles of robotics, robot swarms consist of a large number of robots, each generally small in size, that all come together and work as one single entity to accomplish a certain task. Each robot communicates with its partners in the swarm, gathering and transmitting data, which help to determine how each robot should be behaving. The advantages of having this may not seem obvious at first, because it would seem to make more sense to have one centralized brain power that controlled all these individual robots. However, the programming that goes behind these robot swarms can be shockingly simple. This use of multiple robots, each with a decision making algorithm, is an example of collective artificial intelligence.
The first ever robot swarm was createD in the lab of Radhika Nagpal, the Fred Kavli Professor of Computer Science at the Harvard School of Engineering and Applied Sciences (SEAS) and a core faculty member at the Wyss Institute for Biologically Inspired Engineering at Harvard University” (Perry). Essentially these robots had to come together and created shapes.
The impressive aspect behind these robots was that there was no centralized computer, and each robot was acting based on some communicated information and their own algorithm. In order to complete the task, each robot figured out where it is in relation to the center of the swarm, and it moved along the edge accordingly to complete the shape. This is surprisingly primitive behavior. These kinds of behavior are prevalent among social organisms. For example, “By linking together,” a large colony of army ants “can form rafts and bridges to cross difficult terrain” (Perry). Ants on their own do not necessarily have high intelligence, but when put together in colonies, their collective intelligences can amount to a lot (Perry).
Deriving technology from nature is not something new – the Black Bird airplane, Velcro, Gecko Feet Adhesives are just a few examples. Now, however, instead of simply using the mechanics that nature has to provide, engineers are mimicking the way nature exhibits intelligence, more specifically collective intelligence. In the same way cells of an organism can come together, each performing their own function, to create a living entity, these robots, when put together, can be made to do very versatile tasks.
Termites are a good example of using collective intelligence to accomplish complex tasks. These insects can “build towering mounds several meters high,” with complex architecture featuring “gardens, nurseries, and a bewildering network of tunnels” (Werfel). The main point to be taken away from this is that these termites are working independently, without any sorts of “central supervisor.”
Justin Werfel, in his paper Collective Construction with Robot Swarms, describes the three main obstacles in designing effective robot swarms – “localization, communication, and manipulation”. With a good design in mind, making each of these processes as efficient as possible would result in an effective robot swarm.
Construction is a very compelling use of robot swarms. A lot of man power goes into constructing buildings and it generally takes months to build something. If this could be replaced by robots, the process could become efficient and quicker. If implemented well, this would dramatically reduce the economies of scale when it comes to construction. In the future, designing cityscapes and constructing them would be feasible in a much shorter time than is imaginable now.
Of course the use of swarm robots has only begun and is in its primitive stage, however, as most other things, it should advance exponentially in the coming years. A swarm of nano-bots could be used in a medicinal way, to fend of viruses or bacteria in our body. A version of construction robots could also be useful for relief efforts in places struck by natural disasters – they could come together and get rid of debris, possibly help people out of unsafe places, or simply transport food and aid. Moreover, if self-driving cars become part of our future, then the algorithms that go behind collective artificial intelligence would be very applicable to the cars, where each car acts as an individual among a swarm of other cars. Even without self-driving cars, collective artificial intelligence could be used to control traffic congestion by having distributed sensors around roads, and programming collective algorithms into traffic lights.
Technology has always seemed to advance exponentially, and with the rise of artificial intelligence and robotics, we are certainly nearing a dramatic breakthrough. With AI getting smarter and nano-bots getting smaller, it is imperative that we learn how to harness its abilities properly, and if we do, it can be extremely powerful.
a) Perry, Caroline. “The 1,000-robot Swarm.” Harvard Gazette. Web. 30 July 2015.
Link – http://news.harvard.edu/gazette/story/2014/08/the-1000-robot-swarm/
b) Werfel, Justin. “Collective Construction with Robot Swarms.” Harvard. Web. 30 July 2015.
Link – http://www.eecs.harvard.edu/~jkwerfel/morpheng.pdf