Imagine a situation where you are in your garage and doing some manual work..
e.g., cutting or hammering a piece of wood. All of a sudden, your friend enters the garage and approaches you. To make sure that your friend’s safety is not at risk because of your movements, you have two ways to proceed: you could either stop moving and act like a statue (which would look really weird), or increase your level of attention and change the way you work based on the distance and the actions of your friend.
Believe it or not, most of the industrial robots of today do it the first way! Anytime a worker steps in a robot working area, they stop moving, which usually causes serious delays in the production. For this reason, if you visit highly automated industrial workstations, you will see hundreds of “DO NOT ENTER” signs around to ensure your safety and the production speed at the same time.
To change this trend and make it more like the second approach, a new generation of robots called collaborative robots have been created. The introduction of collaborative robotic technologies opened a new horizon of automation opportunities for agile manufacturing, healthcare, and even disaster response. This was due to their main distinctive features that promoted safety when working with their human counterparts, and enabled adaptation to their unknown surroundings. Nevertheless, in contrast with the increasing public expectation over the past ten years, still not many collaborative robots are integrated in industrial or assistive settings that co-exist or collaborate with humans like the second scenario.
Why is that and how my laboratory’s research aims to address some of the key scientific/technical roadblocks that hampered the expected progress?
Well, first of all, we (and many other colleagues active in robotics research) see humans and robots as a unit. This unit acts like the two humans that change their behaviors continuously and adaptively to make sure that everyone is safe and the work is done with high quality and speed. Obviously, these two conflict and we need to implement some strategies to handle the “limits” all the time and on-the-fly.
Another important aspect is to increase robot autonomy and the decision making capacity, by connecting robot perception (seeing and feeling) to a set of actions. This can improve not only robot adaptation to the surrounding humans and robots, but also to the variations coming from the task and the environment. The last one is commonly defined as “flexibility”, which you might have heard it several times in the context of industry 4.0. In short, it’s all about the ability to change to the varying needs, e.g., of the customers, with minimum set up times and changes in production units. A simple example is that, if you order two personalized cabinets to a skilled carpenter, you don’t need to re-program him/her for the two orders. He/she knows how to change his/her moves to produce the two orders the way you want them. In fact, we want robots to act like this skilled carpenter!
The last, but among the most important topics that my laboratory has introduced within the industrial context is the concept of ergonomic human-robot collaboration. Here, the robots have another superior skill: they can make you work in better ergonomics. This decision was due to a continuous increase of the work-related musculoskeletal disorders in industrial societies, causing hundreds of billions of Euros in annual losses.
The implementation of this vision on such a multi-disciplinary level have required (and will continue to require) multiple funding sources from national and international organizations. In particular, the support I have generously received from the European Research Council (ERC) with the project Ergo-Lean, and from the EU funding programme for research and innovation (Horizon 2020) with project SOPHIA, have permitted me and my lab to bring several of these ideas to life and eventually to higher technology readiness levels (TRL).
The socio-economic impact of our work targeting human-robot collaboration in industry has been essential in keeping the sense-of-purpose of humans in the automation processes, as the humans are not replaced from the workplaces. Instead, they can work in healthier conditions and continuously look for the areas of improvement. This will have a profound impact on the attractiveness of factory work to young workforce, which is due to the creation of high-tech working environments that offer attractive job opportunities to the next generation of workers, engineers, doctorate holders, and researchers. As a result, the problem of labor gap can be addressed, and the industry can become more competitive.
