Photography by Kanishiotu.
Dilemmas in Digitisation
It’s a few hours before dawn on a crisp clear morning. The car park lies in low valley and, even though a frost hadn’t been predicted by the local forecast, the cold air coming off the surrounded hills has become trapped, chilling the road surface to below zero and forming a slippery layer of ice. Yet despite the unexpected nature of the frost, the GRITIT RoboGrit is awake, loaded up with salt and at work – already alerted by sensors embedded in the road surface that had detected the dropping ground temperatures. The robot systematically sweeps back and forth across the parking bays spreading grit, only briefly diverting from the optimised route it follows with centimetre accuracy to avoid the occasional car that’s been left charging overnight.
This isn’t quite science fiction: With some refinement, this is all possible using today’s technology. In fact, RoboGrit isn’t a fictional character either. Not quite. This winter we’ll be starting our first trials of an experimental prototype on real client sites. Looking like a hi-tech miniature dune buggy, GRITIT RoboGrit (the inevitable nickname has stuck) isn’t actually clearing that much ice these days. It is however an important testbed for technologies that include navigation, collision detection and spreading systems. Crucially, many of these technologies are already tried and tested from other industries - drawing on established GPS navigation tools and robotic systems that are already in use in warehouses and factories. Meanwhile, the machine vision systems and batteries being developed for automated and electrified transportation are falling in cost and becoming increasingly available.
As with virtually every other new technology today, RoboGrit is also being designed as part of a wider network of connected, data-driven system. It belongs to the Internet of Things, where smart machines are able to make better decisions with ever greater autonomy and efficiencies. Consequently, work on a gritting robot actually follows our earlier investments into a bespoke system that integrates forecast data to plan gritting runs, and in parallel with ongoing work to develop smart sensors to transmit surface temperatures from sites. Again, both of those developments applied established technologies to the challenge of taking on bad winter weather.
While you're unlikely to see the automation of gritting any time soon, you can see that a surprising amount of progress has already been made. However, technology is only part of the puzzle that needs to be solved. Famously, the author Isaac Asimov invented the "Three Laws of Robotics" - a set of principles by which an imagined future society decided how robots could safely function alongside people. We can say right now with 100% certainly that it’s premature to worry about a robot gritter uprising, but even so it’s still important to think ahead. Look at the challenges being presented by autonomous vehicles, where the human dilemmas and potential legal pitfalls are being cautiously explored. To see this in action, consider The Moral Machine, a brilliant online resource from the Massachusetts Institute of Technology where you can roleplay scenarios relating to the "moral decisions made by machine intelligence, such as self-driving cars". For example, in the face of an inevitable crash, how can and should an autonomous machine choose the lesser of two evils between killing its two passengers or five pedestrians? This grisly illustration, shows the sort of heavy lifting that human operators can’t easily delegate. It’s a key reason why today’s Telsas, BMWs and Volvos that can technically drive themselves on motorways but still require you to keep a hand on the wheel.
In the facilities management context, robots such as automated gritters exist in a lawless frontier where the challenges are still being understood and the rules are as yet unwritten. The same transition can be seen in other areas of modern life being changed by technology: The phrase "unexpected item in bagging area" haunts every visit to a supermarket and shows the friction that exists where messy human life meets dumb, inflexible machine logic. While we can put up with a certain amount of chaos from an automated retail experience, in safety critical contexts such as transport, manufacturing, and (soon), snow and ice removal, AI will still need a human hand on the wheel.
For this reason, it is also premature to see the introduction of robotics into sectors like facilities management as a threat to jobs. Instead, data and robotics will for the foreseeable future be much like previous industrial revolutions - with technology serving as force multipliers that make human workers more productive and efficient. For example, it's most likely that your first sighting of an automatic gritter will be as it clears larger areas like car parks while a supervising human worker tackles trickier areas like stairs and paths while monitoring the robot’s performance and carrying out any necessary maintenance. Just like many other industries disrupted by technology, gritting will continue to be a human industry but workers will increasingly swap manual labour for a more sophisticated set of skills. And it will be worth it: Teams using technology will be able to clear more locations each night to an even better, safer standard. The result will be lower risk to those working and visiting a site, with a reduction in costs and liability exposure for owners and managers.
Today, sophisticated bespoke technology is already helping our people deliver an award winning service that’s been tested under the toughest conditions. But there’s always room for improvement: We’re looking forward to welcoming you to the team, RoboGrit.