Human-robot collaboration (HRC) in shared workspaces is on the verge of a breakthrough. The main concerns are improving ergonomics, making work processes more flexible, increasing efficiency and optimizing logistics, handling and loading processes. Two main directions can currently be identified in the context of HRC:
On the one hand, lightweight robotics is booming. A number of companies have started application demonstrations and tests in the pre-series environment and are currently gathering experience. At the same time, new suppliers of robot technology are entering the market and creating intense competition. Assembly assistance, the 3C market (Computer, Communications, Consumer) and the Asian economic area are driving the current debate. As a result, all suppliers are working on optimizing weight, costs and performance.

On the other hand, the shortage of skilled workers and demographic trends mean that there is an increasing focus on ergonomic relief for workers. This applies in particular to demanding assembly environments, for example in gear manufacturing. In the collaborative partial automation of process chains, humans continue to take on the central joining and guidance tasks. At the same time, they are supported by robots that pick up the loads that occur. This primarily involves handling weights in the medium load range. This requires collaborative end-of-arm tools that are capable of handling different loads in the range of up to 2 kg in direct interaction with humans. However, common solution strategies for inherently safe HRC grippers - such as harmoniously shaped protective covers without sharp corners and edges and limiting the gripping force to 140 N - reach their limits in such scenarios: Firstly, they do not have the appropriate gripping force, and secondly, they lack flexibility in terms of stroke.

This is precisely where the 'Co-act EGL-C' large stroke gripper, which Schunk is presenting at the Hannover Messe, comes in. It achieves gripping forces of up to 450 N and combines this with a stroke of 42.5 mm per finger. The intelligent 24 V power pack thus opens up new scope for collaborative handling of workpiece weights of up to 2.25 kg. By the end of 2019, it will be launched in series production with a DGUV certificate for collaborative operation and interfaces for HRC robots from Kuka, Yaskawa, Fanuc, Universal Robots and Nachi, among others.

Combined force and displacement measurement

In order to comply with the biomechanical limit values defined in ISO/TS 15066 despite the high gripping force, the gripper is equipped with combined force and displacement measurement: Force measuring jaws integrated into the base jaws and incremental encoders permanently monitor the respective gripping force as well as the position of the gripper fingers. The gripping procedure on the gripper is divided into several phases: The gripping force is limited to 30 N up to a theoretical distance of 4 mm from the taught workpiece - significantly less than the thickness of a finger. If a collision occurs during this approach phase, for example with the operator's hand, the gripper immediately comes to a safe stop without the risk of injury.

Powerful gripping in human-robot collaboration: the Co-act EGL-C is the first large lifting gripper for collaborative applications.

© Schunk

Only in the second phase - i.e. when the workpiece distance is <4 mm - do the fingers close with a freely definable maximum force of up to 450 N. If the system detects yielding in this closing phase, for example because a workpiece is gripped that is too small and the operator is about to remove it by hand, this movement also stops automatically. The same applies if the expected workpiece dimensions are exceeded by 2 mm, for example because no part is present. Finally, in the third phase, the gripper detects whether the part is securely gripped and activates the integrated gripping force maintenance by applying the brake. This means that the gripped part cannot be lost even in the event of an emergency stop. In addition, no re-referencing is required in the event of a power failure.

The safety architecture on which the new generation of grippers is based was developed in close cooperation with the DGUV. It is based on several integrated processor cores that monitor each other and at the same time safely control the gripping process via the force and displacement sensors used. This ultimately makes it possible to realize gripping forces beyond the 140 N limit that has been common up to now.

In addition, the Co-act EGL-C concept was designed from the outset with a high level of practicality in mind. The gripper can be controlled and regulated via Profinet, Ethercat, Ethernet/IP, Modbus/TCP or TCP/IP. A commissioning wizard simplifies programming. In addition, a diagnostic interface enables access to the gripper's most important process and status data during operation. In order to make cooperation with the operator fluid and intuitive, the gripper is equipped with LED lighting in traffic light colors, which can be used to signal the current status of the module.

Next step: Autonomous gripping

In the context of human-robot collaboration, the use of artificial intelligence (AI) is also becoming increasingly important. The first applications of cognitive intelligence in the gripper environment have already been implemented in conjunction with cameras, enabling intuitive training by workers and independent completion of gripping tasks by the robot.

Thanks to artificial intelligence, this 5-fingered hand is able to identify any object in any position and autonomously develop and apply appropriate gripping strategies.

© Schunk

An initial use case at Schunk, which uses machine learning approaches for workpiece and gripping process classification, shows how pluggable building blocks can be combined as required and presented to a lightweight robot in any arrangement on a work surface for removal. In combination with 2D or 3D cameras, the self-learning system achieves a rapid increase in access reliability after just a few learning cycles: with every grip, the gripper learns how the workpiece can be successfully picked up and transported. After just a few training rounds, the network classifies how to deal with the value pool of workpieces and the resulting possible combinations. In doing so, the gripper relies on learned empirical values on how to pick up and transport the workpiece. The intelligent performance of the algorithm is that future combinations and arrangements of the workpieces can be classified independently after just a short training period. This enables the system to handle parts independently and according to the situation. By continuously adapting the algorithms using AI methods, it is possible to identify previously unrecognized correlations and further refine the handling process.

Interaction - the be-all and end-all

The topic of human-robot collaboration (HRC) is challenging manufacturers and users alike to rethink their approach. Professor Dr. Markus Glück, Managing Director Research & Development at Schunk, takes a stand.

Mr. Glück, what trends are currently shaping the HRC market?
Prof. Dr. Markus Glück: We are currently seeing manufacturers of robot arms and the corresponding control technology developing into full-range suppliers of robot systems and handling processes. Following the example of Universal Robots' UR+ platform, highly interesting ecosystems and exchange forums are emerging. In addition to the robot arms, web-based sales and user platforms will offer customized components that can be ordered online and put into operation in a simple but reliable manner in the sense of efficient plug & work.