Home / Blog / Post-Pandemic Scenarios – XXVII – Robots and Company 2

Post-Pandemic Scenarios – XXVII – Robots and Company 2

Kilobot Robot Swarm Image credit: asuscreative

Robots Swarms

Robots are groing in number and they are learning to cooperate. As the number of cooperative robots involved in a specific task grows, their coordination becomes more and more tricky. However, we have examples in nature where seemingly low-intelligence entities manage to cooperate achieving impressive result as a whole. Bees and ants are obvious examples and they are not alone. Aggregation of independent living entities often generate complex structures (let’s not forget humans that manage -often without directions- to achieve impressive results that would seem to derive from some organised control Think about walking in a crowded street: you are usually not bumping on other people as both you and them take diversive action to avoid collision and that is done without any “rule” nor coordination).

Scientists have discovered that group behaviour can derive from some very simple rules that are shared among the group and enacted independently by each entity in the group. Take a look at the clip, showing the choreography by a flock of starlings. There is not orchestra director, each bird is on its own, yet together they create an amazing performance.

Now researchers are working to use similar approaches to enable cooperation among hundreds of robots. An example is Kilobots, shown in the picture. They have been developed by researchers at Harvard at a cost of 14$ per unit. They are so cheap because each one of them has stripped down functionality but they can act as a single body and all together perform complex activity. This is (was) just a sort of possibility demonstrator, but the idea is catching up.

As reported by the World Economic Forum, Walmart filed (2018) a patent for autonomous robot bees to serve as pollinators (this may turn out to become a crucial technology as bees are being decimated in some parts of the world and bees are a fundamental cog in the food chain).

The creation of a robot-swarm needs to take into account a distributed decision making, based on individual decisions, including:

  • consensus: a general algorithms supports individual decisions that converge onto a specific goal. Decisions may differ (like one robot decides to go left, another to go right) but they all result in a convergence towards the goal;
  • tasks allocation is distributed to maximise the global efficiency and does not require a single control point. Allocation results from local awareness (like I am overloaded you are not….);
  • collective fault detection: each robot in the swarm is able to detect if the behaviour of the neighbouring ones can be trusted (in other words if they are functioning as expected) and automatically exclude those perceived as “aberrant” from its set of relationships;
  • continuous resynchronisation leading to a dynamic local change of behaviour that creates ripple and reaches all corners of the swarm (this is exactly what happens in flocks, look at the clip: local changes result in a wave of change affecting the whole and in turns creating a feedback on the local environment).

For an interesting overview/roadmap of robot-swarm evolution take a look at this document.

Smart dust

Take a robot and squeeze it to the sub-millimetre scale (even to the micro scale, and to exaggerate to the nano scale), then multiply it by thousands or tens of thousands: what you get is smart dust, tiny speckle of silicon that embeds sensing, micro-processing and local communications capabilities. Their very existence (and mutual position) correspond to a state of the swarm.

Smart dust was imagined over twenty years ago and their properties demonstrated in the lab (Berkeley). At that time they were “big”, like a cubic cm, but the idea was to have them much much smaller, the size of a dust particle (hence the name). They have already shrunk to the sub cubic millimetre and in the second part of this decade they will get even smaller. The bigger ones, in the mm range are also known as “motes”.

According to the FTI’s report they will find application in several fields, including sensing the brain activity and interacting with peripheral nerves in bioelectronic applications (neural dust). Sensing in various fields is likely to remain their main area of application, flanking the IoTs and extending their application. Military, agriculture, factory automation, healthcare are seen as the main areas of application in this decade.

About Roberto Saracco

Roberto Saracco fell in love with technology and its implications long time ago. His background is in math and computer science. Until April 2017 he led the EIT Digital Italian Node and then was head of the Industrial Doctoral School of EIT Digital up to September 2018. Previously, up to December 2011 he was the Director of the Telecom Italia Future Centre in Venice, looking at the interplay of technology evolution, economics and society. At the turn of the century he led a World Bank-Infodev project to stimulate entrepreneurship in Latin America. He is a senior member of IEEE where he leads the Industry Advisory Board within the Future Directions Committee and co-chairs the Digital Reality Initiative. He teaches a Master course on Technology Forecasting and Market impact at the University of Trento. He has published over 100 papers in journals and magazines and 14 books.