How is Colruyt Group Technics' maintenance model evolving?
“We are implementing a process to monitor and manage the entire life cycle of our assets", Department Head of Machinery Brecht Kestemont says. “We want to move from 'maintenance' towards 'asset management' and strive for a reliable installation at the best possible cost", Reliability & Maintenance Engineer Joeri Debontridder adds.
From study to maintenance
Our Machinery department examines the study, design and realisation of production machines, installations and automations, but also acts as Single Point Of Contact for all maintenance and repair assistance of these technical installations. It does so, on the one hand, for the various logistics distribution centres (Dassenveld, Ghislenghien, Ollignies, Stroppen, Mechelen, Londerzeel and Laekebeek), and, on the other, for Fine Food's production sites. “At Fine Food Coffee we roast and package coffee and at Fine Food Wine (Ghislenghien) we package wine in bottles and party boxes.”
“Fine Food Meat cuts and packs meat, and Fine Food Meat 2 in Halle is Belgium's largest cold cuts production site with about 700 machines", Kestemont says. The production site was proclaimed 'Factory of the Future' in 2019. Cheese cutting and packaging at Fine Food Cheese (Dassenveld) is fully automated, as is the production of all kinds of salads such as tuna, crab and seasonal salads (filling, sealing, packaging and crating) at Fine Food Salads (Stroppen).
Including partner in the story
In 2019, we started using the Overall Equipment Effectiveness improvement tool (OEE) at Fine Food. That project was completed in June 2021. “The tool helps to provide more insight into the performance of the machinery and the bottlenecks in the production process. It shows the actual effectiveness of the machine versus the ideal machine.” Including the partner in this story is imperative. “Data are essential. In addition to OEE data, we also monitor the data generated by the machine itself (counters that record operating hours, temperature and vibration sensors, pressure sensors, etc.) for our critical installations. We however also rely on work reports and registered data regarding interventions by our technicians.
From maintenance to asset management
In shifting our focus to asset management, we called on Mainnovation's expertise to further refine work methods and techniques and develop medium-term and longer-term actions. “For example, we fine-tune FMEA (Failure Mode and Effect Analyses), determine stock values, and we draw up Long-Term Asset Planning (LTAP). In doing so, we look at machines at ten to fifteen years, their need for maintenance, upgrades etc." Twenty employees took the course which helped us look at things objectively and determine where we can improve. Our analyses are now even more data-driven. We map out the actual situation and we can determine more precisely what benefit our actions can mean for the customer", Kestemont says. "In the past, we often relied on our 'gut feeling' with regard to the operation of a machine. That does not always correspond with the data. We removed the subjectivity from our criticality ranking (what is critical and what is less?).”
ERP provides streamlining
Meanwhile, we have installed a new ERP. “The existing system already registered a quarter century. We opted for Planon which offers more functionality," Debontridder says. The ERP tool streamlines the entire maintenance activities, from the work order to the task ID, the intervention on the line and the machine. But it also keeps track of machine costs, enabling trend analysis and answers to the why question to help determine strategy. “With the new tool, for example, structured reports on maintenance and repairs can be completed automatically.” The digital forms use so-called SORA codes: symptom, origin, remedy and action. A thesaurus of keywords has been agreed upon which provides clarity. The unambiguity brings about a very special acceleration.
Maintenance and support
The Machinery Department provides maintenance and support for these productions. The entire machinery is in the hands of 240 people, of which 200 technicians operate on the floor, divided into fourteen technical teams with nineteen reliability & maintenance engineers (RMEs). The technical team and the day-to-day operational activities are managed by a team manager, focusing on reliable and available machinery.
Each technical team works 24/7 in two shifts, with separate night and weekend shifts. Profiles sought are electromechanics, industrial sciences engineers. Engineering pur sang. In addition to short-term operational planning, the service is involved in medium- and longer-term tactical and strategic planning.
“We implement more and more data-driven maintenance", Kestemont says. Measurements on components are already being performed today. Just think of temperature and vibration measurements and noise measurements to determine the potential failure (PF) interval. "We however want to make our machines and maintenance even smarter by using sensors and components that generate data." We for instance initiated a POC on some of our installations to proceed with vibration analyses by means of smart sensors. The advantage of such system is that we can monitor our installations in real time. In addition, we installed sensors on our highest energy-consuming devices to gain insights into our power, compressed air and water consumption. In this way, we have our installations run according to need, thus reducing our energy consumption.
Also for our standard maintenance tasks such as lubricating, we recently started using "smart lubrication." Smart lubrication involves the use of intelligent lubricating nipples and an intelligent lubrication pump. Both systems are built in order to allow two-way communication. A colour code enables the technician to see in real time whether the lubrication was done properly.
The data exchanged is sent to a central database through the cloud, where it serves as input to a dashboard that tracks lubrications and component condition. Any follow-up actions can be organised from there on.
Internet of Things
In a next phase, we want to evolve further in terms of IoT and artificial intelligence, to analyse causes of failures through correlations in the data. In that way, we are moving towards machine learning. We want to take everything to a higher level to automatically raise alarms and plan and mobilise people in an action-oriented manner before machines and production chains fail: predictive maintenance at its best.”