HVM & ARAM: Human-Robot Collaboration Validation and Automated Risk Assessment
Collaborative robots make it possible to design flexible and agile assembly lines without physical barriers between humans and robots, thus reducing required factory spaces and improving productivity. However, the implementation of Human Robot Collaboration strategies requires several design iterations to guarantee a minimum level of risk to the operator while maintaining a satisfactory level of productivity. The risk evaluation process is time consuming. Complying with normative constraints such as ISO 15066 while maintaining ergonomics and productivity can be complex.
The SHERLOCK consortium was motivated by the following assessments:
- Manufacturing companies have difficulties understanding and complying with the safety standards of ISO/TS 15066 (management of shocks, crushing risks between the operator and the robot, etc.), which are very technical and sometimes misunderstood or misinterpreted by workstation designers.
- Traditional simulation tools do not provide strategies to optimize safety (choice of hardware, sensors, safety parameters, cell layout)
- Operators often do not participate proactively in the design of the working system, their practical knowledge is not considered which limits the acceptance of the solution.
The outcome of SHERLOCK consists in two exploitable results: HVM and ARAM. HVM stands for Human-Robot Collaboration validation module and is a software tool while ARAM (Automated Risk Assessment) is a methodology.
The HVM module (commercial name XRTwin) is a simulation-based tool helping to interpret and understand the safety standards of ISO/TS 15066. It follows the design process of the HRC applications and is intended to be used by all participants in the workstation design (robotics engineers, robot integrators, ergonomists, safety experts, operators, decision makers).
HVM provides design tools using Extended Reality technologies to gather all actors around the workstation design. It allows to identify inadequate design of safety strategy, inadequate positioning of sensing devices, operator body regions exposed to mechanical hazards, excessive efforts caused by non-ergonomic design or heavy loads.
Detailed description of HVM features can be found here.
The ARAM tool, integrated into the HVM module, is a support tool for conduction the risk assessments based on ISO-12100 methodology within a VR environment. It allows to produce more accurate and realistic the risk estimation having as result a better risk assessment, including the possibility to detect and estimate potential hazards caused by physical fatigue.
From the safety and health technicians who conducts the risk assessment of machinery and HRC cells in design stage, the tool is bringing the following benefits front traditional methodologies:
- to conduct an accurate risk assessment. Since the virtual representation is a realistic simulation is possible to detect and outline potential hazard that would be no evident performing a traditional risk assessment based in diagrams and printings.
- the risk assessment module is providing the user with a higher level of certainty when identifying hazards (more concrete, when, where, which body parts, etc.), whereas the traditional methods at design stage are usually based on assumptions.
- introduces the possibility for the safety expert to detect in design stage those nonconformities while making easier and cost-effective to implement the risk reductions measures to the machine manufacturer, just re-editing the virtual representation
HVM and ARAM have proved to be very useful in the design of the safety strategy for the SHERLOCK use cases. We have shown that, using CAD based immersive simulation tools, the user (designer or integrator of cobotic technology) will be able to perform a risk analysis within the virtual environment and identify any non-conformities at the early stages of the design phase. The accuracy of the dynamic simulation and impact estimations is validated through on field real on a collision measurement tool (PILZ Prms®). As a result, these outcomes of SHERLOCK are extensible to most Human Robot Collaboration use cases if 3D data are available.