Guided Safety – Guided safety design, configuration and execution of small part assembly process with collaborative robots
Manufacturing end-user: Denso Automotive Deutschland GmbH – Production Engineering
The Business Sector
Denso is one of the world's largest automotive suppliers. For decades, this industry has been at the forefront of the latest automation technologies, which are indispensable for competitiveness. While the degree of automation in some production stages is almost 100 percent, assembly is still largely carried out by hand. With the help of technological innovations, many improvements in terms of efficiency and economy can still be achieved.
DENSO is a leading supplier of advanced automotive technology, systems and components for major automakers. Its main products are components for automotive electronics and automotive mechatronics. The parent company is headquartered in Kariya, Japan. Established in 1949, the company has today about 170,000 employees organized in 200 group companies. The annual revenue is US$ 48.1 bn. Of this, 9 percent is invested in research and development activities. Denso holds 38,000 patents worldwide.
Automating assembly processes is challenging due to various requirements. These include short cycle times, small batch sizes and many product variants. Therefore, and for the realization of flexible production systems in the sense of industry 4.0, production systems must be increasingly flexible, reconfigurable and easily adaptable. This will not work with rigid automation solutions as they have been widely used so far.
Instead, new key technologies such as human-robot cooperation will be needed to enable Denso to respond to the increasing demand for its products and the shortage of qualified specialists. Hybrid workstations at which humans and robots work in teams according to their abilities allow scalable and flexible (partial) automation of joining processes. In order to realise these jobs economically, some hurdles still have to be overcome. These include a reliable risk assessment on the one hand, and the safe design of the workstation and the safe configuration of the safety functions on the other. In the “Challenge”, the project partners address the design and configuration, since the first point has already been realized in the EU project LIAA. The goal is to expand the HORSE framework with technologies that The goal is to extend the HORSE framework with technologies that simplify, accelerate, track and automate the realization of a secure production process as far as possible. Denso shall be enabled to quickly and reliably commission hybrid assembly workstations and thus increase productivity.
In a first step, the shift manager logs into Camunda to configure the line:
- assign workers to manual tasks
- assign robot programs to automated tasks
A special plug-in was developed to allow her/him to do so in an intuitive way, by basically selecting the appropriate option from a drop-list.
Product job initialization
In a first step, the worker at station 1 needs to import the production order by scanning in a QR-Code encoding the variant to be produced and the lot size. The variant information determines which branch to follow in some local tasks, see local tasks in station 3 for example, whereas the lot size information determines the number of process instances to launch.
Operations in station 1
Once the QR-Code is scanned, the worker in station 1 immediately receives a message to start the assembly of the corresponding variant. He is asked to press “Completed” once he is finished with this task. Next, the business engine determines the state of the buffer immediately after station 1. If it is “empty”, then the pallet is moved from station 1 to the buffer and the next pallet is loaded in the station accompanied with a new message containing the assembly instruction and so forth.
Operations in station 2
The pallet will remain in the buffer between station 1 and 2 until station 2 signals that it is empty. At that moment, two processes are supposed to happen: a) The conveyer loads the pallet into station 2 and b) the robot moves to the “Home” position (preparation task). Next, the robot executes its assigned assembly program. In this station, the assembly tasks are independent of the variant. The business engine determines the state of the buffer immediately after station 2. If it is “empty”, then the pallet is moved from station 2 to the buffer and the next pallet is loaded in the station.
Operations in station 3
The pallet will remain in the buffer between station 2 and 3 until station 3 signals that it is empty. At that moment, two processes are supposed to happen: a) The conveyer loads the pallet into station 3 and b) the robot moves to the “Home” position and picks up the 1st screw to be inserted (preparation task).Next, the robot executes the screwing task, i.e. screw in the first screw, pick up 2nd screw and screw it in. In case of variant 3, an additional assembly task is performed, where the robot picks up an assembly part and inserts it in the base plate. The business engine determines the state of the buffer immediately after station 3. If it is “empty”, then the pallet is moved from station 3 to the buffer and the next pallet is loaded in the station. There are two additional buffers on the backtrack to station 1.
Operations in station 1 / completion of job
Once the pallet with the finished assembly reaches station 1, the worker is instructed to unload the finished product. A small script checks whether the lot size of the production job has been completed. If not, the worker is instructed to proceed with the assembly of a new iteration. If the lot size is completed, then the pallet is unloaded into the buffer and so forth until all pallets are unloaded.
In case of errors or failures during the local execution of the robot tasks, the local executer (FlexBE or Drag&Bot) are configured to report a “NOK” result. A watchdog is implemented that monitors this result.
In case of “NOK”, it triggers a manual recovery action, where it is up to the worker to decide among three options:
- Retry, i.e. he resets the pallet configuration and triggers a retry of the robotized assembly operations.
- Continue, i.e. he completes the robotized assembly task manually, and then triggers the unloading of the pallet into the buffer.
- Fail, i.e. the error is so sever, that the production job has to be cancelled and the maintenance crew has to be notified.
Of course, if station 2 breaks down, there is still the option to bypass it by reassigning tasks to station 3. A bypass can only be triggered by the shift manager. He will also have to select different robot programs for execution in station 3.