The development of robotic systems and their applications has rapidly increased in recent years. The challenge today is, e.g. that the robot performs different successive subtasks to achieve one or more complicated tasks similar to humans, and at the same time, to have the ability to cooperate with humans safely and efficiently on different levels of interaction [1]. In advanced collaborated robotic applications, using only one kind of feedback is sometimes insufficient to achieve the desired goals and to cover all the safety requirements.
In order to obtain more information on the work environment, it is preferable to use different kinds of sensors such as vision sensors, force sensors, laser sensors, tactile sensors, etc. However, in terms of control, more sensors mean more possibilities for the structure of the control system.
Although most of the algorithms for Human Robot Interaction consider safety procedures, the industry is still missing solutions that prioritise productivity and efficiency during human interaction with industrial robots
In general, most of the developed control algorithms for Human Robot Interaction (HRI) have taken into account the safety procedures during a defined type of HRI. However, if the safest industrial robot has already been developed but it is uneconomical (too slow, stops very often because of stringent safety procedures everywhere, large safety distance between human and robot, huge cycle time, etc.), it will never be used in industry.
3D Zone-Based Robot Control
According to the standard ISO/TS 15066, in speed and separation monitoring (SSM) operation mode, the human robot shared workspace should be divided into different zones, e.g. safe, warning, danger zones, etc. Usually, this division is based on the distance between human and robot, the type of the Human Robot Interaction, the recurrence and on the accessibility of the human near the robot.
These factors define the robot velocity and the required sensors for monitoring the workspace in every zone. In general, these rules ensure the human safety during interaction with industrial robots. However, productivity, efficiency and diversity of the shared tasks are not really considered.
In Sharework EU-funded project, the Fraunhofer IWU team proposes a strategy for modelling any shared workspace between humans and robots. The focus lies on establishing a generic methodology for dynamic risk assessment and developing one part of the proposed system called 3D zone-based robot control, which is built using 3D advanced selection matrices.
In every zone, each robot‘s subspace could be controlled by vision, force, position etc., depending on the required tasks, the safety requirements and risk analysis. Using the proposed system, the industrial robots can perform various shared tasks with the human, including different control algorithms and safety procedures.
During Sharework project, a novel strategy for modelling a shared workspace between humans and robots based on dynamic risk assessment and 3D advanced selection matrices will be validated
Various parameters in every zone can be easily configured, e.g. a) dimensional/ dynamic characteristics, b) priority of zones in situations where they overlap, c) allowed objects/events, d) required feedback for robot control etc.
The main advantages are:
- Increasing system availability, due to “less breakdown on account of safety procedures“.
- Economising the sensors and safety procedures without endangering humans. It takes into account the standards, e.g. ISO 10218-1/2:2011 and ISO/TS 15066:2016, concerning HRI applications. However, the workspace will not be exhausted with various sensors everywhere.
- Optimising the synchronization between human and robot cycle time. In other words, increasing the robot’s productivity of the robot during complex tasks by establishing an automatic relation between human position, robot position, workspace and the control loop of every robot subspace.
- Improving the HR workspace flexibility when any modification or upgrading is required.
FPGA-Based Parallelisation
The rapid evolution of sensor technology in robotics has paved the way for the appearance of collaborative robots that can cooperate and assist workers in industrial applications. However, the workspace should be fully equipped with different conventional safety sensors for avoiding any injuries. The most commonly used conventional sensors for safety in robot cells are laser and vision systems.
More sensors means more processing and communication time. Here the challenge lies in how to speed up the time of data processing and the communication between sensor, robot control and safety controller. Sharework will exploit the powerful and capabilities of the FPGA technology to minimize the safety distance between human and robot and to optimize the facility’s layout.
Safety as a condition for greater productivity and flexibility
Safety is not our target. It is a condition, which should be always fulfilled. Our target is to increase the productivity and flexibility and because of that the developed models in Sharework will open new horizons to establish not only safe but also flexible and efficient Human Robot Interaction applications, starting from cobots and ending-up with heavy-duty systems.
Dr.-Ing. Mohamad Bdiwi
Dr. Bdiwihas been working at the Fraunhofer IWU in Chemnitz since 2013. He leads the Department of Robotics. Dr. Bdiwi has received his PhD degree from the Technical University of Chemnitz, Germany, Department of Robotic Systems. He has published more than 30 international scientific papers and journal articles. Furthermore, he has been leading many national and international projects and managing the CTC-Collaborative Technology “Cluster for Human-robot interaction and robotics applications”.
Fraunhofer IWU Team in Sharework:
Aquib Rashid, Paul Eichler, Lena Winkler and Mohamad Bdiwi
References
[1] A new strategy for ensuring human safety during various levels of interaction with industrial robots Mohamad Bdiwi *, Marko Pfeifer, Andreas Sterzing, CIRP Annals, Volume 66, Issue 1, 2017, Pages 453-456
