Afstudeeropdracht: AI-ML Parameter Advising Veghel • Actemium

AI-ML Parameter Advising

Many of our customers are concerned about using cloud-based solutions due to their policies that restrict data from leaving the physical plant. To address this concern, we plan to develop an on-premise (local) machine learning (ML) tool using open-source software. This ML tool will help operators optimize the production process by providing advice on the best parameters to use. 

For example, consider a heating process controlled by a PID controller. A PID controller is a type of feedback control system used to maintain a desired output, such as controlling a steam valve. It uses three parameters: P (Proportional), I (Integral), and D (Derivative). To achieve the most energy-efficient and stable temperature, these parameters need to be set correctly. 

Optimizing these parameters can be challenging. First, it takes time to see the effects of any changes made to the parameters. Second, there are many variables to consider, such as humidity, ambient temperature, product temperature, and the speed at which the product moves through the oven. 

Project Objectives 

1. Develop a Local ML Tool: Build a machine learning tool using open-source software that can operate locally without relying on cloud services. 
2. Parameter Optimization: Program the ML tool to recommend the optimal P, I, and D parameters for the PID controller. 
3. Expand Functionality: Enhance the tool to include additional inputs such as product speed, product temperature, and environment temperature. 



Project Challenges 

1. Selecting Open-Source Software: Identify and choose the appropriate open-source software for developing the ML tool. 
2. Programming the ML Tool: Develop and program the ML tool to provide accurate recommendations for the PID controller parameters. 
3. Testing and Validation: Test the tool in various scenarios to ensure its recommendations improve the production process. 

At Actemium, you have the space to trust in your own qualities. Of course, with an open mind for the knowledge and experience of your colleagues. In practice, this means: 

• A pleasant, informal working atmosphere and tailored guidance from enthusiastic professionals. 
• Plenty of room for your innovative ideas. 
• An internship or graduation opportunity with a lot of variety within a network that is constantly evolving. 
• (Opportunities for) growth after completing your assignment. 
• Flexible working hours 
• Market-based compensation 

• You are almost finished with your study 
• You are enthusiastic and being passionate of technology 
• You can work on your assignment independently on our location in Veghel 
• You want to keep going forward. Standing still is no option. You want to keep developing yourself, within your assignment but as well on personal field 
• You can communicate in a nice and professional way and want to lift the conversation to a next level 
• You are able to advice, even if not asked 

Actemium kijkt vooruit. En gaat vooruit. Op alle fronten. Met onze diensten brengen we onze opdrachtgevers en daarmee de wereld verder. Samen schakelen we naar een duurzame toekomst. Wie er bij ons ook vooruitgaat? Onze 1250 collega’s verspreid over 25 zelfstandige businessunits. Dankzij de vrijheid om initiatieven te ontplooien. Door opdrachten die onze expertise uitdagen en topcollega’s die elkaar bij staan. Het vormt onze basis om op te groeien. En iedere collega krijgt oneindig veel ontwikkelmogelijkheden om dat te doen. Actemium is onderdeel van VINCI Energies, een van de grootste internationale spelers in de industrie. Kijk maar eens op  wat we allemaal doen!