Robots will soon be a necessity but they won’t take all our jobs

by Stelian Coros, 14 Jan 2019

In the future, robots will be ubiquitous but they will not take everyone’s jobs. Advances in the field should pave the way to a future in which robots and humans can work together with ease.

There are four key trends in robotics at present.

1. Adaptability

The recent surge in Artificial Intelligence (AI) and related technologies is a boon for robotics. Indeed, with applications ranging from personal assistance to therapy, and from social companionship to search-and-rescue, the socioeconomic promise of this field is now unquestionable. Whether to augment human capabilities in the workplace, to help with chores around our homes, or to increase the effectiveness of disaster-response operations, future generations of robots will need human-level mobility and manipulation skills to perform an increasingly diverse array of tasks.

2. Personalisation

Thanks to new technology and continued innovation, the process of creating robots will become significantly more accessible than it is today. Intuitive interfaces and software will enable the design of entirely customized robots, perfectly adapted to different tasks and to the needs and preferences of users.

By lowering the barriers to entry in the field of robotics, a much broader segment of the population will have an opportunity to engage in activities that are currently restricted to domain experts.

This has the potential to influence the field as profoundly as open source software has shaped the IT landscape. The idea here is for human designers to create bespoke robots with minimal effort in a user-friendly software platform. In short, engineering meets AI, helping people to upskill in order to adapt to the fast-changing world of work that is quickly becoming a defining challenge of our generation.

3. On-demand delivery

This trend takes the ability to customize robots even further. The production and assembly of 3D-printed parts not only enables on the spot custom-designed robots, but also self-assembly on-site. It is like LEGO for scientists. Researchers wanting to deploy a robot for a task in the field would be able to design it according to their criteria and actually generate its components themselves, before assembling and using it immediately.

Customers interested in companion robots could, in a similar way, select their preferences, complete their order on one day, and enjoy their purchase the next day – all without breaking the bank.

4. Choreography

AI is also teaching robots how to move, thus automating the learning process by which they grasp the rules of the environment in which they evolve and how to avoid or react to obstacles.

Robots have traditionally moved in a slow, cumbersome fashion. Novel trajectory optimization allows users to generate walking, rolling, gliding, and even skating motions – again, as required or preferred. These motions emerge based on the components used to design each individual robot.

This feature, in turn, allows robot motions and behaviours with stable, physically-viable motions that are realistic and compelling. Using this new design methodology, a diverse array of hybrid legged or wheeled mobile robots can not only be imagined, but also realised.

What is next in robotics?

Emerging digital fabrication technologies and novel materials are paving the way to increasingly lifelike robots. The crucial trend will be dealing with complexity.

The disruptive power of additive manufacturing (AM) technologies comes from their ability to fabricate designs of unmatched geometric complexity using a continually expanding range of materials. More remarkable still, AM allows the mechanical properties of printed structures to be engineered based on high-level specifications, by digitally controlling the geometric arrangement and composition of individual base materials.

By mastering the unprecedented fabrication capabilities enabled by AM, over the next decade we will be able to create synthetic structures that approach the mechanical sophistication of their biological counterparts.

This will lead to new breeds of robots that will resemble biomechanical systems in both form and function. These robots will abandon traditional designs that rely exclusively on stiff materials and actuators in favour of compliant alternatives that will redefine their performance, efficiency, robustness and safety characteristics.

While robots often belong to the realm of science fiction, technology is facilitating their integration into society. The trends discussed above show that we are closer than we think to robots becoming ubiquitous.

At the same time, these advances will all help to create more meaningful human-robot interactions, avoiding the dreaded scenario whereby robots take over everyone’s jobs. On the contrary, they have the potential to provide much needed assistance that is both complementary and beneficial.

Professor Stelian Coros will present his robots at the ETH.