First of all, we need to be clear about the definition of an autonomous system; there is no standard which defines properly what the levels of automation are. So Naïo provides a pragmatic approach to get the right vocabulary that will lead us to do better risk assessment and better categorization about the required performances for Naïo’s system, depending on the level of autonomy that we want to achieve.
As soon as a function is able to take command of the motion of the machine, and thus that can decide to go one way or another, we consider it as an autonomous function.
At level 3, there is the robot, defined as assisted autonomy: a machine which is able to work alone and take decisions by itself. At this level, a task is done autonomously without any perception system. The operator is needed nearby the machine to take decisions in case of obstacles in order to react properly: an autonomous work supervised by a human.
At level 4, a perception system is introduced on the machine which is able to sense and send signals to the operator. At this level, the machine is able to detect and to make restrictive decisions. But humans still need to make complex decisions.
Level 5 is about full autonomy: the machine is able to do all the job. Autonomous work, make decisions if there is an obstacle, and a supervisory system which is able to take complex decisions. There is still a human in the loop, who is aware about what is actually being done by the machine.
So if you know what level of autonomy depending on the risk assessment, you are able to define the proper level of performances needed for your critical system.
There are two major risks in the field, but the main one is hurting someone with the robot. If someone enters the danger area, you need to shut down the power. Then, the operator has to come and launch the machine again. Naïo’s approach is to have two areas around the machine:
- The first area is the immediate vicinity of the machine where there can have an immediate danger
- The other area around the area where the operator stands and can make a decision
Cédric Séguineau, Head of HSQE at naïo Technologies, explains the different safety approaches.