Robotic co-worker

+ We appear to be at a tipping point of technology that will enable collaborative robots to be used much more widely in the workplace. Robots of the future will be able to interact safely and learn from human co-workers.

I believe we appear to be at the tipping point of technology that will enable collaborative robots to be used much more widely in the workplace. Robots of the future will be able to interact safely and co-operate with human co-workers, as well as learn from them. 

I am no expert on robotics and artificial intelligence but, as a futurist, I am increasingly drawn to this complex, rapidly evolving and very exciting field. The implications will be far-reaching, and we may have to adapt our lives to work alongside increasingly sophisticated robots (many of which won’t be humanoid in form, but could be a collection of disembodied sensors).

Robots will replace or augment not only unskilled, routine jobs, but also many highly paid and highly skilled jobs. These could include doctors, journalists and financial traders (automated algorithms are already responsible for high volumes of financial transactions).

In a recent report on the future of factories, we look at the increasing use of collaborative robots in manufacturing. Factories have long used industrial robots for tasks that involve heavy lifting or repetitive jobs that require speed and precision. However, these robots have been too unintelligent and dangerous to work alongside humans, who tend to perform more delicate final assembly jobs or tasks that require flexibility.

This is changing. Take Rethink Robotics’ Sawyer robot, which can be taught to perform tasks by human co-workers who have no programming expertise. A human can physically guide Saywer’s arm through part of an activity and Sawyer can then infer the rest of the task.

One of the most interesting developments is the ability of robots to learn to perform complex tasks without being reprogrammed, by building knowledge through trial and error. They learn from experience and can adapt their behaviour to improve upon a task. 

For example, a robot from the Berkeley Robot Learning Lab called BRETT (which stands for Berkeley Robot for the Elimination of Tedious Tasks) uses deep learning to complete tasks without input from humans. Using trial and error it has learnt to assemble a basic toy plane and to place a Lego brick in the correct position. 

In the next decade, we could see robots learning complex tasks from scratch. These robots will learn in a similar way to humans, through consuming information, demonstrations by others, and trial and error.

This ability to learn as well as interact safely with humans will have implications far beyond the factory floor. One day, we could see robots taking over manual labour tasks such as painting walls, cooking meals, repairing roads, folding laundry or walking the dog. But it is a mistake to think only manual labour will be affected; we all have elements of our jobs that are predictable and subject to automation.

While the reality of your employer replacing you with a robot may still be some way off, we need to start considering whether our education and professional training systems are fit for the robotic age. As robots become more flexible and responsive, human workers will need to develop new skills and take on more creative or supervisory roles, or they will become redundant. Ultimately, humans will need to possess more flexible skill sets than their robotic co-workers.

As futurist Alvin Toffler has noted: “The illiterate of the 21st century will not be those who cannot read and write, but those who cannot learn, unlearn, and relearn.”