Swarm robots need not be the stuff of nightmares, says Tom Longstaff. In fact, they offer huge potential to the public good
Remember New Year’s Eve last year? What an innocent time that was. Few, if any of us, had paid much heed to early reports of a new virus emerging in China, and while there was certainly much uncertainty around us, there was no inkling of the seismic changes to come.
That particular New Year, though, was also notable for the now customary shots of cities near and far welcoming the arrival of 2020. Perhaps reflecting their futuristic image, some cities, such as Shanghai for example, chose to augment the traditional fireworks with hundreds of drones – and the results were spectacular.
Hundreds of robot drones, moving in orchestrated synchrony to form complex patterns and images in the sky? Sounds incredible, and one could easily be forgiven for thinking that the dawn of robotic swarms had arrived. As impressive as it was, however, this amazing display was not actually a swarm in the true sense of the word. To explain why, let’s start with a simple definition.
Introducing swarm robotics
Swarm robotics, in essence, is the study of how a swarm of relatively simple physically embodied robots can be constructed to collectively accomplish tasks that are beyond the capabilities of a single robot – hence the New Year’s display. Akin to swarms of insects in many ways, a large number of robots can perform complex tasks more efficiently than one alone.
Using multiple robots to complete a task has many advantages over a single robot. Robotic swarms can withstand the loss or damage of members, can adapt their behaviour to undertake a variety of tasks, and can scale in size to meet the changing demands of a given task.
The most remarkable thing – and the key aspect that differentiates them from basic multi-robot systems – is that robot swarms are entirely decentralised; there is no leader and no centralised control. It is perhaps unsurprising, then, that swarm robotics and Artificial Intelligence are closely linked.
Back to our New Year’s spectacle. As impressive as it was, it was centrally controlled by a computer. A true swarm would be self-organising; able to execute complex pattern formation without any central control or direction.
Nevertheless, we must dream big: the potential of robotic swarms far exceeds an annual display come every December 31st. Such swarms can be put to work in all manner of settings and given all manner of tasks, and lead to better results than might otherwise have been achieved.
Take search and rescue, for example. Such missions are, sadly, a familiar sight around the world, and can range from the hunt for a missing toddler to the aftermath of a natural disaster affecting thousands over multiple hectares of land. Technology has already revolutionised the way such activities take place – think thermal imaging cameras, for example – but robot swarms can take this a step further and potentially prevent huge loss of life in the process.
Their ability to traverse an unknown and dangerous environment, overcoming potential obstacles such as fire, floods and chemical substances, is based on factors such as the robustness of a swarm to individual unit failure, as well as the scalability of emergent behaviours to swarms of different sizes. The bottom line is that the results from a swarm are greater than the total work by individual units – collective power, in other words.
And of course, search and rescue is by no means the only area where robot swarms might be deployed. From military missions to crop harvesting, environmental monitoring, remote exploration, medical diagnostics and intervention, unmanned space exploration and infrastructure maintenance, the opportunity is vast.
Follow the science
So now what? We’re still some way from robot swarms being successfully deployed into real world situations, at least at scale. To realise swarm robotics we must first conquer some design and implementation challenges. Chief among these are the design of behavioural algorithms and building swarms of robots that can handle the physical and computational burden.
If these and other challenges can be overcome, swarm robotics has huge potential. But everything we know from the scale of the progress so far, not to mention the research that is underway in laboratories around the world, is that swarm robotics are fast morphing from theory to reality.
About the author
Tom Longstaff is an Operational Cyber Technical Advisor at BAE Systems Applied Intelligence
Explore Government Insights
Stay up to date with the latest thinking, trends, technologies and projects from our Government teamFind out more
- Creating the Navy of the future. The Royal Navy is in the midst of a concerted effort to exploit and deploy advanced data analytics and Artificial Intelligence. Hannah Green explains why it’s full speed ahead
- Shaping a future that works for all. Charles Newhouse says we must balance excitement about AI’s potential with support for those it may leave in its wake
- A moving story. Battle-scarred commuters are a testament to infrastructure buckling under the strain of carrying an ever-growing number of passengers – but is hope at hand? Miriam Howe explores the impact of technology on the transport systems of the future
- 5 ways emerging technologies can help tackle Coronavirus. As the Covid19 pandemic continues to envelop the world, Roberto Desimone considers how emerging technologies can help resolve both the current crisis, as well as prevent future outbreaks from taking root
- Turning synthetic technologies into real impact. Libby Vallance-Bull explains why synthetic environments may be key to unlocking advantage in future highly unpredictable conflict