Micro-robots Poised for Big Growth
David Cotriss | October 02, 2014A relatively new type of robot is making its way into the industrial automation and electronics manufacturing sectors. Often referred to as micro-robots, these devices are as small as 1 millimeter across and can build structures and form shapes by working together without human intervention.
After slow growth the past two years, it appears the global market for industrial automation equipment (IAE) is turning a corner. IHS analysts expect the market to reach $185.3 billion in revenue in 2014, up from $173 billion in 2013. IHS predicts the sector to hit $225 billion in revenue in 2017.
Multiple drivers are behind this predicted industry growth. Convergence, the idea of melding manufacturing with new technology to improve efficiency, and an aging engineering workforce with relatively few replacements will combine to necessitate greater automation. IHS also predicts that tools such as semi-autonomous robots will play a major role in spurring growth and eventually leading to automatic factories.
Another factor that will contribute to the demand for greater automation is size. As consumer electronics devices like smartphones continue to demand precision manufacturing methods, robots capable of performing delicate and detail-oriented tasks will be required.
The electronics and semiconductor sector is a small but growing segment of the industrial robot market. According to the Robotics Industries Association, the sector grew by 89% in terms of units sold and by 56% in sales revenue in 2012, the most recent year for which figures are available.
With that as background, research teams at SRI International and at Harvard University have developed biomimetic micro-robots. These minuscule robots act much like colonies of ants do in nature, working individually as part of a collective whole to accomplish tasks.
On the industrial side, SRI’s project is a U.S. government investment as part of the U.S. Department of Defense's Advanced Research Projects Agency's (DARPA’s) Open Manufacturing Program. The program aims to “lower the cost and speed the delivery of high-quality manufactured goods with predictable performance” by “creating a manufacturing framework that captures factory-floor and materials processing variability and integrates probabilistic computational tools, informatics systems and rapid qualification approaches.” SRI’s micro-robots comply because they are autonomous and flexible in what they can assemble (electronic components and even small structures).
Harvard’s version of micro-robots measure 33mm in diameter. The “Kilobot” moves via vibrating motors and communicates with other Kilobots up to 7 cm away using infrared light. By working together to overcome their individual shortcomings – a single Kilobot offers limited capabilities and is error-prone – the Kilobots are able to form shapes. Eventually, the Harvard team would like its robots to be able to self-assemble into 3D objects.
SRI and Harvard’s micro-robot swarms operate at different ends of the command spectrum. SRI’s robots operate using a centralized control. Harvard’s approach is de-centralized: four seed robots receive commands from a human operator, after which they work with the rest of the swarm to accomplish their goal.
Justin Werfel, research scientist at Harvard’s Wyss Institute for Biologically Inspired Engineering, says that SRI’s centralized approach may help boost efficiency. “The controller can optimize exactly what all the robots do,” he says, “so that the job gets done in the least possible time, with the shortest distances traveled.”
Alternatively, Werfel says the two main advantages to a decentralized system like Harvard’s are scalability and robustness. “You can add more agents to a decentralized system without necessarily needing to change how they’re programmed or how many the system can handle, whereas a centralized controller may be a bottleneck.”
Werfel says he believes SRI’s version of micro-robots is likely most useful with small-scale projects (at least currently) since they require circuit boards to run on. Harvard’s Kilobots, on the other hand, may make more sense in scenarios where “large-scale control over details of the environment is more challenging, and individual robots can more easily have their own sensors and controllers,” Werfel says.
SRI and Harvard are not the only institutions carrying out experiments and research using swarms of micro-robots. Teams at the University of Sheffield and University of Lincoln, in the UK, and the New Jersey Institute of Technology in the U.S., are working on similar swarm-based micro-robots, albeit using smaller numbers of robots.
The micro-robot movement is nascent and the market still developing. But as the needs for more precise manufacturing and automation rise, products such as SRI’s micro-robots and Harvard’s Kilobots may prosper, especially if projections for the industrial robot and manufacturing industries hold true. Buoyed by mobile device and similar markets’ need for components, growing demand for new manufacturing processes and investment from organizations such as DARPA, micro-robots may help shape the future of manufacturing.