Remarkable improvements of production capacity – A heavy mould turned into an AGV
Trygons designs and manufactures large and complex parts from reinforced plastics outside of Athens, Greece. Its 4000-square-meter manufacturing facilities also allow cost-effective mass production.
Despite using, and developing, advanced manufacturing technologies, Trygons relies heavily on manual work when it comes to intra-factory logistics. When manufacturing roofs for trucks, for example, moulds weighing 500 kg are lifted and carried by workers from workstation to workstation.
Making one piece requires fours stops in the workshop and takes altogether several days. First a gelcoat is painted onto the mould, then the cut fiberglass is placed inside the mould and the resin is infused and cured in a vacuum bag. After demoulding, the mould is cleaned and prepared for the next production cycle.
Besides long set-up and process times, manual logistics and paper-based coordination also cause physical and mental strain, safety risks and errors.
The L4MS experiment offered Trygons an opportunity to find out how to benefit from using an AGV (Automated Guided Vehicle) to automate the moving of moulds.
Present and future benefits from 3D simulation
Trygons was joined by specialists at the National Technical University of Athens (NTUA) and its Institute of Communication and Computer Systems (ICCS). They started by creating a digital twin of the manufacturing process. This was done with the 3D simulator Visual Components provided by OPIL.
“3D simulation of the complete manufacturing process was very useful for us, and Trygons already considers using it with further development projects,” says Senior Researcher, Dr George Avdikos from ICCS.
OPIL platform to connect operators and the AGV
ICCS specialists developed the solution on the OPIL platform to which an AGV was connected as well as the workstations. The mould was actually turned into an AGV by attaching wheels, a microcomputer and a specific printed circuit board to it.
The AGV starts moving when it receives a command from OPIL sent by the operator at a workstation. It follows a white tape attached to the floor with the help of a magnetic sensor. When it reaches the required workstation, it stays there until the work at the station is done. Then the operator presses a button on a tablet and OPIL sends the command to the AGV to start moving. This is repeated at four workstations in the production unit.
Besides using existing OPIL modules, also a new OPIL module was developed.
“The new module SAN will further extend OPIL capacities, as it can support sensors’ drivers independently of the user’s programming language,” George Avdikos says.
22% increase in production capacity
When testing the solution, the following Key Performance Indicators were confirmed:
- Moulding takt time was reduced from 45 to 35 minutes per cycle.
- Production output capability was increased by 22%: from 14 pieces to 18 pieces per day, if AGVs were used in all five production lines.
- Faults in final products were reduced so that time spend on finishing and repairing was reduced by 15%: from 120 minutes per piece to 108 minutes.
- Overall quality and versatility were improved.
“We manage to validate a proof of concept: communication between the AGV and the OPIL platform through a low latency communication link. This technology can also be applied with minimal customisation to other companies when automating their logistics or production. We have discussed this with customers from the automotive and marine industries,” George Avdikos says.
The cost of the project was 42 500 EUR including materials and work involved in design, development and testing, and production downtime. It took approximately 2.5 months to design the solution and 2 months to test it.
OPIL modules used in the Trygons solution by ICCS
- OPIL Sensor Agent Node (SAN): to connect various sensors to OPIL and provide data about them to Orion Context Broker
- OPIL Human Machine Interface (HMI): to define logistics tasks
- OPIL Task Planner module: to guide the Robot Agent Node (RAN) of the AGV
- OPIL Robot Agent Node (RAN): to send control commands to the AGV
- 3D simulation tool Visual Components: to create a digital twin
Partners in the L4MS experiment:
- Factory: Trygons
- System integration and technology provider : ICCS, Institute of Communication and Computer Systems at National Technical University of Athens (NTUA)
- Software integration: General Department at National & Kapodistrian University of Athens (NKUA)