The problem
Aalberts Surface Technologies machines, assembles, and repairs precision parts, mostly for the power generation industry. Its assembly floor handles fuel nozzle and cover assemblies, both new-make and repair, along with their sub-components, all in the same space. Because the product mix keeps changing, no single layout fit all of it, and parts and operators were covering far more ground than the work required. That showed up as longer cycle times, labor spent walking instead of building, and, in the worst cases, mislabeled parts that customers rejected.
Our five-person team ran a full DMAIC project on the floor, working with the operations manager, the manufacturing manager, and the assembly crew.
Measuring the waste
Distance alone didn’t capture the problem. Carrying a 76 lb nozzle body across the floor is not the same as walking over to grab a tool, and a part built every week matters more than one built once a year. So we built a primary metric in ft-lb-vol: distance traveled × part weight × that part’s share of annual volume. A secondary metric tracked operator trips without a part, like fetching a nitrogen hose or walking to a bench to key in a part number.
A Pareto analysis of three years of order history narrowed the focus to the five part families that make up about 80% of volume. For each one, we mapped the process, walked the floor, and drew spaghetti diagrams of every move.
Finding the root causes
Cause-and-effect diagrams and why-why analysis surfaced about 20 causes, which we scored in a cause-and-effect matrix against part and operator movement. Four tied for the top: the floor layout didn’t follow the process flow, typical batch sizes exceeded workbench capacity, there was no standard batch size to design carts around, and there were no designated buffer areas. Parts sat in walkways, carts blocked paths, and nozzle bodies had to loop around the cover assembly area to reach inspection.
The fix
We went through three proposed layouts before reworking the design with the production team, since they knew constraints we couldn’t see from a drawing. The final layout gives new nozzle builds a single-piece flow path, frees a separate line for repairs, opens a direct route to labeling and inspection, and adds staging areas for batch parts. Order history showed nozzle bodies usually ship in batches of 14 or more, so we also proposed carts sized to the real batch.
Results
The target was a 20% reduction in the part movement metric. The final layout cut it by 40.7%, with operator movement down 48.6% and net travel down 15.5%. The biggest gains came on the two highest-volume nozzle families, at 60% and 81%.
Making it stick
A new layout drifts back without controls, so we set up 5S color coding and floor taping for every zone, labeled staging areas, and mobile shelving to keep parts near the work. We also recommended a shop-floor schedule display and barcode scanning to replace paper tracking.