Sometime in the future we will be working with robots. They can do complex tasks already and in some factories they are already working hand in hand with human coworkers. But how can it be done safely?
Scientists at KTH Royal Institute of Technology created a new system, which provides robots with a whole new level of context awareness.
Obviously, the problem is that robots might make moves that do not take into account their surroundings. A quick move from a robot may be enough to cause a significant injury for someone. This should be avoided through sensors that help robots judge distance between themselves and humans. However, there have been tragic accidents when robots killed their human coworkers.
Scientists now say that understanding the distance is not enough – robots need to understand the pose of the human and even predict the next pose. These kind of abilities could bring robots to a new level of context awareness, needed for safe interaction with their work environment.
This new system employs Artificial Intelligence (AI) solution, which requires less computational power and smaller datasets than traditional machine learning methods. This could make the system easier to implement. However, for now this technology is out ahead of today’s International Organization for Standards (ISO) requirements for collaborative robot safety. In other words, it is so new and so advanced that international institutions do not take its abilities into account yet.
Current ISO standards require robot coworkers to slow down when a human is near and stop entirely if someone is close enough to interfere with its movement. This is a low level of context awareness and it does interfere with productiveness. If we are to achieve a true collaboration between robots and people, they need to be able to work close to each other without interfering work. Slowing down just because someone is near, is not acceptable. That is why robots in this KTH Royal Institute of Technology system actually avoid obstacles. They are not limited to proximity – they predict where the person is going to move and how.
Hongyi Liu, one of the researchers in this project, said: “This is safety not just from the technical point of view in avoiding collisions, but being able to recognize the context of the assembly line. This gives an additional layer of safety.”
Future factories will be filled with robots. Even more so than they are now. But humans will continue to work there as well. Safety barriers restrict space, which is at a premium in some cases. And current ISO safety standards have a negative impact on productivity. We need to change that quickly. Hopefully, studies like this will help create technologies that would encourage policy makers to change the legislation and improve productivity and safety when robots are working among humans.
Source: KTH Royal Institute of Technology