Situation awareness in industrial robotics

The aim of this work is to investigate the impact of new robot movements (e.g. slow and repetitive) on human situational awareness. Man and robot would share information with each other and this information would be taken into account in each other’s decision making scheme.

  • A new method is proposed for the assessment of risks in robotics. Numerous accidents are caused by situation awareness errors. The idea is to use Endsley’s model and the 8 situation awareness demons to categorize industrial accidents.
  • Three important situation awareness demons are involved in most industrial robotic accidents: out-of-the-loop, errant mental model and attentional tunneling. A Bayesian network is proposed to determine risks probabilities.
  • In addition, specific recommendations are made to improve situation awareness and reduce the risks, especially providing a living mode to help the operator understanding that the robot is not stopped. For instance, a “breathing” movement added to the robot when it is stationary but not stopped would inform the interacting human that the robot is active and therefore potentially risky.

Experiments are currently carried out to demonstrate the relevance of the approach. The idea of the experiments is to show that the robot motion can be considered as a signal that would inform the person about the state of the robot. Thus, this signal would enhance the person’s situational awareness and help him make the right decisions to accomplish his task. If the participant sees the robot performing a breathing movement, he must understand that the robot is in an error situation and still active. Therefore, knowing the procedure, he has to follow it to rectify the error and avoid an accident.


Comments are closed.