In the safety technology of collaborative robots, standards differentiate between
various collision types, the identification and differentiation of which are essential
for ensuring safe operation. The objective of this paper is to develop and test a
mechanism actuated by artificial muscle to examine the detection of these profiles
in different collision scenarios. The ISO 15066 standard distinguishes between two
types of collisions: quasi-static and transient. Using a simplified model, experiments
were conducted to evaluate whether sensors could identify collision types accurately.
The results demonstrate the feasibility of identifying collision types through sensor
data. The findings have the potential to enhance the safety of collaborative robots.
