References PC mask
Share to linkedin Twitter
DAF

DAF

DAF Trucks - The added value of Plant Simulation 

 
The challenges:
  • Simulation of a complex production line;
  • Assessment of proposed investments;
  • Optimization of series sizes and operating rules. 
 
The approach:
  • Use of Plant Simulation software to create a realistic and integrated model;
  • Customized training for DAF Trucks staff;
  • Most of the modeling done by people employed by DAF Trucks.
 
The advantages:
  • A highly accurate and realistic model of a complex production line;
  • Investment in additional production capacity proved unnecessary;
  • Extension of material handling systems to make heavy manual handling tasks less strenuous;
  • Greatly improved knowledge of the production system.
 
In 2006 DAF Trucks needed to scale up its engine production to meet increasing demand. The production engineers at the company’s engine production plant realized that new methods were needed to model, analyze and optimize production processes. Positive experience with a supplier led to the decision to use discrete event simulation to optimize camshaft production. If the project proved successful, discrete event simulation would become the main tool to support the upscaling of the whole production process. cards PLM Solutions BV rose to the challenge with Plant Simulation software from the Tecnomatix portfolio.
 
Growing complexity
 
Growing demand posed a considerable challenge for the Production Engineering Department. The first step in upscaling production was to use simulation as part of a pilot project. The project began with one simple question: should the engine production plant use discrete event simulation to model, analyze and optimize the production systems? To answer this question a pilot project was set up for the camshaft production line. The line involved a number of issues which made it an interesting case for simulation. Niek Jansen, a Process Engineer and Project Manager at DAF Trucks: “We wanted to gain a clear understanding of the possibilities of the software. At the same time we also had very specific questions, such as is the capacity of the line essentially adequate or do we need to make further investments to keep up with demand? Are the series sizes optimal? Would managing the process in a different way improve performance? Does the increasing complexity of the engine production plant mean that more powerful tools are needed to analyze and optimize the production processes?” These questions formed an excellent starting point for the simulation project.
 
First experiences
 
cards PLM Solutions BV provided (on-the-job) training for two production engineers using Tecnomatix training material. The majority of the analysis and modeling would be done by people employed by DAF Trucks. Niek Jansen, Project Manager: “It was very important to us to build the model ourselves. This was the only way we could really assess the effectiveness of the software.” Since neither of the engineers had worked with Plant Simulation before the project was also a litmus test of the user-friendliness of the software.
 
The training was provided at the head office of cards PLM Solutions BV. While the production engineers were being trained information was gathered about the production of camshafts. Every simulation project begins with meticulous analysis of the system being modeled. The characteristics of the production process and the flows within the production process must be known. This preliminary analysis is necessary in order to be able to build a valid model. Virtually every operational process includes anomalous routines and characteristics that need to be identified in order to be able to build a realistic model. At the same time meticulous analysis shows which details are relevant and which are not. Gert Nomden, a consultant at cards PLM Solutions BV explains: “Every minute invested in preliminary analysis pays off in terms of smarter decisions during the model-building phase. There is also an interesting spin-off in that the questions raised during the preliminary analysis phase generate a great deal of process knowledge. This phase also ensures that the people on the work floor are more involved in improvement projects.”
 
The production process consisted of processes such as milling, tempering, grinding and washing. Most of these processes were automated. Material handling processes were also automated with the aid of a portal robot. The product flow consisted of different types of camshafts, which required a switch-over from one type of production to another. The frames in which the camshafts were buffered formed another flow. Different types of data had to be collected, such as processing times, production switch-over times, fault patterns and other such data. Historical production data was also gathered as a benchmark that could be used to validate the accuracy of the model. Project manager Niek Jansen also stresses the importance of thorough preliminary analysis: “We were unable to answer some of the questions that the cards PLM Solutions consultant asked about the production process. But we needed to be able to answer these questions in order to build a sound model. Observing the operation of the line and talking to the people on the work floor were an essential part of the project.” The process flows and layout were mapped in various diagrams together with the operation of the processes performed by the portal robot.
 
Initial results
 
The first parts of the model were built shortly after the first training sessions. The initial components were developed in line with the modular and hierarchical concept of Plant Simulation. A first rough version of the model produced surprising results right from the start. Niek Jansen explains: “Our first model simply incorporated provisional versions of the main components of the production line. Yet even this very early model was surprisingly accurate in terms of parameters such as throughput and capacity utilization rate.” This first model was gradually elaborated and increasing levels of detail were added to make it more realistic and more accurate. This was achieved by incorporating the actual operating logic of the portal robot, the different types of camshafts and the supply of camshafts, among other things. It was at this stage that a serious problem arose.
 
Project manager Niek Jansen: “I was suddenly called in to deal with an unexpected technical problem at the engine production plant. With the prospect of a significant delay in the project we decided to outsource the modeling to cards PLM Solutions BV. The transfer of the project was accomplished very smoothly.” The modeling that still had to be completed at that stage consisted of a section that involved mainly manual work and the supply of camshafts from the tempering process. Because the project was a pilot project, the model had to be as realistic as possible. Special decision routines had to be built in and the animation had to look realistic. Gert Nomden: “In this particular case the animation had to mimic reality perfectly. Most of the operators were not familiar with simulation. In reality the portal robot also moved vertically, so the robot also had to be seen to do this in the model. The first impression created by the model had to be right. Any discrepancy would undermine the credibility of the project!” Once the final details had been implemented in the model it was time to start the validation phase.
 
Proof of the pudding
 
The validation phase is designed to verify that the model can answer the right questions with sufficient accuracy. From the start of the project the operators gathered detailed information about the output of each shift. The ERP system provided production data over a longer period. The engineers conducted a series of time studies to determine the capacity utilization of machines and operators. The simulation model was then fed with historical planning data and the output of the model was compared with the actual outcome in the past. The output of the model and the actual outcome were extremely similar, the production quantities differed by just 0.5%. Foreman Theo Opdam made a bold contribution to the project: “I decided to carry out an experiment with the existing production line. We set up the camshaft line and the model so they were both in an identical starting position. Then we let them both run for a while and compared the results. Again the results were virtually identical. This confirmed that even a process as complex as a camshaft line with a portal robot and operators can be accurately modeled with simulation!” With the reliability of the model having been established, the project team could start to answer the other questions that needed to be addressed by the project.
 
Surprising advantages
 
The maximum feasible output of the camshaft line was determined by evaluating a series of demand scenarios. It turned out that the existing line configuration was actually adequate to meet future demand. Investment in extra production capacity, which had already been budgeted for, ultimately proved unnecessary. The simulation results also showed that the utilization rates of the manual section of the line were very high. This was a key factor in the decision to make a relatively small investment in automated material handling systems that would make the heaviest tasks less strenuous for the operators. The next aspect that needed to be assessed was the series sizes. The performance of the production line was sufficiently robust to accommodate all series-size scenarios. The final aspect was the operation of the portal robot. While other operational parameters did not really affect the performance of the line, the analysis of this aspect of the production process yielded several critical insights regarding the key role of the portal robot for the production line as a whole.
All in all Niek Jansen was delighted with the results: “It would have been impossible to gain all these insights without discrete event simulation. I was extremely pleased that we were able to make more efficient and effective investments. By the time we completed the project we had recouped our investment in the project more than ten times over! Armed with these results Niek Jansen had little trouble convincing the senior management of the added value of simulation." Marcel Hoedeman, Production Engineering Manager at the DAF Trucks engine production plant
 
Expanding success
 
The progress and results of the simulation study were observed with interest by many people at the engine production plant. During the project the simulation team received a whole host of questions about possible applications. The questions pertained to ideas for various production lines at different stages of the process life cycle. Some were conceptual, such as what would be the result of a radically different line-feed concept? Others focused on detailed engineering and optimization of existing processes, such as what measures do we need to implement to increase the output of the production line? All of these ideas were collected in a simulation portfolio and assessed against certain criteria, such as how urgent is the project? What efforts are required to implement the project? What are the anticipated advantages? Is the project in keeping with the creation of a DAF Trucks objects library?
 
It will come as no surprise that DAF Trucks decided to pursue the successful application of simulation. By the end of the project the company had invested in several Professional and Run time licenses. cards PLM Solutions BV provided training for other production engineers. A simulation team was formed to address all of the simulation questions raised at the engine production plant. Internal procedures were adjusted to determine which questions should be answered with the aid of simulation. Production Engineering Manager Marcel Hoedeman: “With simulation we are able to design optimal processes, and we can do this faster and with less investment. Simulation is now embedded in our design guidelines. We have the right people and the perfect tool to further optimize our production processes!”
 
http://www.daf.com