Johannes Kurth and Marcel Wagner
Human-robot collaboration (HRC) remains a topic that everyone is talking about. There are successful applications. The major breakthrough has not yet occurred, however. One main reason for this is that in the introductory phase, HRC was only integrated into existing systems that were not prepared for HRC, nor suitable for it. The full advantages of HRC can only be exploited in new systems, and, even then, only if new system planning methods are used. This article highlights what changes are necessary and presents an extended planning methodology and planning principles for HRC.
Since the presentation of the first robots with corresponding safety functions for collaborative operation in 2014, HRC solutions have been integrated into existing systems, but the resulting solutions have only rarely been technically and economically convincing. Why is this?
HRC in existing systems
Existing systems (brownfield) were never planned and prepared for human-robot collaboration. On the contrary. Work was consistently divided between either manual workstations or fully-automated robotic stations behind a safety fence, light curtain or scanner. The manual workstations were synchronized in such a way that workers are optimally employed in terms of their abilities and the time required. Two very important aspects:
- Human abilities and degree of utilization
If one were to attempt the technical replication of a human, it would consist roughly of the components illustrated:
A 7-axis robot would be just one of many optimally coordinated components. This shows just how unfair the competition between humans and robots actually is. Does a single robot really stand a chance? Yes, whenever the abilities of humans are not fully utilized.
The second important aspect:
- Utilization of time
In linked assembly workstations, the work is generally designed in such a way that each worker can perform his specific tasks well within the cycle time and without longer pauses occurring. If an existing system is well planned in this respect, it becomes apparent why the advantages of HRC are not fully effective there. Integration of an HRC robot generally only partly relieves the load on the worker. It is rarely possible to make sensible use of the worker’s saved working time, however, as the cycle times of the upstream and downstream workstations are optimized. The workers do not require assistance. Other tasks are often located too far away, which would lead to unproductive walking distances.
However, if one considers not only a single workstation, but is also prepared to redesign several workstations, this opens up a whole new range of possibilities, as the work can now be divided according to new criteria.
HRC in new systems
In the case of HRC in new systems, a whole new set of parameters apply. All planning aspects of HRC can be taken into consideration from the outset. Work can be divided optimally between humans and robots. This means that the full potential of HRC can be exploited. Furthermore, considerably higher investment resources are available than for the conversion of existing plants.
New planning methods for HRC
In the case of a new system, all assembly steps must first be classified. Two questions must always be answered here:
- Firstly, is the process step also suitable for automation under cost-effective conditions?
Not everything that is technically feasible is also costeffective.
- Secondly, can a process step that can be automated also be implemented in combination with HRC?
A welding process, for example, would not be suitable for HRC.
If these two points are answered in the affirmative, a plant concept is drawn up. If this concept was again positively evaluated, the plant detailing begins.
The following aspects should be taken into consideration when designing the system concept:
- Avoiding imitation of the tasks of the human worker
To the robot, many of these human limitations are unknown. It does not need to take ergonomic factors or exhaustion, etc., into account. This opens up entirely different design options.
- Revolution rather than evolution
Previous concepts must be set aside, and processes and work contents need to be rethought.
- Utilization of robot capacity
When assigning work to the robots in a system, it should therefore be ensured that each robot only carries out a single process where possible. By means of a clever arrangement of the individual workstations in the layout, the system can be designed in such a way that the robot takes over tasks in two or more stations and thus utilizes its full capacity once again.
- Task-centered decision for fixedcycle or continuous-flow operation
In very large systems, such as final assembly in an automotive plant, it remains to be decided in which areas the system is to run in fixedcycle or continuous-flow operation. In this context, the sequence of the assembly of specific components should also be reconsidered. If the priority graph allows, assembly tasks can be shifted to preferred zones with fixed-cycle or continuous-flow operation.
These new aspects of HRC-oriented planning initially pose a certain challenge for the planner. The reward for this effort is a highly productive, cost-effective system with ergonomically optimized workstations. Furthermore, it is now also possible to automate processes that previously fell at the hurdle of excessively high costs for automation-friendly in and outfeed of workpieces. This is enabling automation to advance into areas that were previously closed to it.