The shop floor has a daily start meeting, shadow boards for tooling and an agreed cap on work in progress. Twenty meters away, in the engineering office, none of that exists. There, waste is simply called being busy, and being busy is considered part of the job.
That is remarkable, because engineering is a production department too. The product just is not steel but information: models, drawing sets, bills of materials, exports. And wherever things get produced, things get wasted. Translate the seven classic wastes of lean to engineering and you will recognize them within a day: waiting for a release, files traveling back and forth, models polished further than anyone asked for. A toolbox like Thundercad can remove part of that technically, but lean does not start with software. It starts with recognizing what leaks away.
This article walks through the seven wastes, each with its engineering equivalent and a first countermove. It deliberately stays at process level: the five most concrete click-and-wait examples inside Inventor itself are covered in a separate article, more on that below.
Waste in a department that produces information
Lean defines waste as anything that consumes resources without adding value for the customer. For engineering, "the customer" is twofold: the paying customer, and internal consumers such as work preparation, purchasing and the shop floor. Value is everything they need to decide, order and build. The rest, however skillfully executed, is waste.
Two caveats before you start crossing things out. First, not all waste can be avoided: a quotation that does not win, documentation the law requires. Lean calls that necessary waste; you minimize it, you blame nobody for it. Second, waste almost never lives in people but in the process around them. The engineer who spends an afternoon hunting for the right file is not lazy; he works in a system that has made searching normal.
Waiting, transport and overprocessing
1. Waiting: work standing still
A design that sits for three days because two customer questions are open. A drawing package waiting until the project lead has time to release it. A variant stuck because the lead engineer is the only one who knows the calculation model. Waiting is the quietest waste: nobody does anything wrong, and yet weeks go by. The countermove: bundle questions and decisions. One fixed weekly meeting covering all open points across all projects beats ten separate emails that each sit unanswered for days.
2. Transport: files traveling back and forth
In the factory, transport is a forklift moving pallets without adding anything to them. In engineering it is an export emailed to work preparation, saved there, adjusted, sent back and saved again. Every handover is a queue plus a chance of a wrong version. The countermove is not faster emailing but less moving: one source location where consumers fetch the current state themselves, and handovers that consist of a reference instead of a copy.
3. Overprocessing: more polish than requested
The purchased part remodeled down to its thread. The drawing with eight views for a part the supplier machines straight from the 3D model. The tolerance tighter than the saw can hold, which the shop floor therefore silently ignores. Overprocessing feels like craftsmanship, which is exactly why it is stubborn. The countermove: ask per document type who uses it and what they minimally need in it, and dare to leave out the rest.
Recognizing waste is step one, removing it is step two. Thundercad automates the routine work around exporting, BOMs and metadata, so engineering hours go back into design.
Try 30 days freeInventory, motion, defects and overproduction
4. Inventory: half-finished work on the shelf
Work in progress is invisible in engineering: no pallets in the aisle, but twelve half-finished drawings, three variants "we will get back to" and a release that only happens once the whole project is done. That inventory hides problems, exactly as it does in the factory: errors only surface when the entire batch is released at once, at the moment time pressure peaks. The countermove: smaller batches. Release per subassembly as soon as it is done, so work preparation and purchasing can start earlier and errors show up earlier.
5. Motion: searching, clicking, navigating
In the factory, motion is the fitter walking to the warehouse counter three times an hour. In engineering it is hunting for the right file, navigating folder structures and repeating the same click sequence for the tenth time that day. Per action it is nothing; added up over a week it is a full working day. The five biggest examples inside Inventor, each with a remedy, are covered in 5 time-wasters in Autodesk Inventor and how to eliminate them. At process level the question is always the same: how many actions sit between wanting something and having it done?
6. Defects: data errors that land downstream
A wrong quantity in the BOM, a missed change, an export of an old state. The peculiar thing about defects in information is the delay: the error is born in engineering but explodes at purchasing, in the workshop or at the customer, and gets booked there too. So the source stays out of sight and the error repeats itself. The countermove starts with feedback: every data error the shop floor finds goes back to engineering with its suspected cause attached, otherwise nobody learns anything.
7. Overproduction: making what nobody asked for
According to lean the heaviest waste, because it feeds all the others. In engineering: working out variants no order exists for, exporting "just in case" in four formats, weekly lists nobody reads anymore. Everything you produce unasked also has to be managed, updated and searched afterwards, and so it inflates inventory, transport and the odds of defects.
Make it visible: walk one order through
Fighting waste without measuring turns into a yes-no debate. The lightest measurement that works: walk through one completed, representative order, from intake to release.
- List every step the design went through, including handovers and round trips.
- Note two times per step: how long was actually worked on it, and how long the work sat waiting.
- Label every step with one of the seven wastes, or with the label "value".
- Add up the waiting time and put it next to the touch time.
The outcome is almost always uncomfortable: a design that took four weeks turns out to contain only a few days of net work. Do not read that ratio as blame but as a map: where the waiting sits, your biggest waste sits. If you want hours and a yearly picture attached to it, use the worked example in What does repetitive work really cost your engineering team?.
One waste at a time, with an owner
After measuring, pick the waste with the most hours and agree on one countermeasure, with an owner and a review moment after a few weeks. Five measures at once dilute each other; one measure that visibly works creates appetite for the next. Often the first candidate is a routine job that returns every week. The export moment is a grateful example: with Batch Publish a complete drawing package goes to PDF, DWG, DXF and STEP in one run, instead of drawing by drawing through the same dialogs.
Finally, watch out for shifting instead of removing: a department that releases faster by dropping checks moves the waste into defects further down the chain. The seven wastes are a connected set; the measurement above tells you whether the total drops or merely changes boxes.
Frequently asked questions
Is lean not primarily a shop floor thing?
Lean comes from series production, but the logic applies wherever work passes through hands in steps, so also where the product is information. Better still: much shop floor waste, like waiting for dimensions or rework after a data error, originates upstream in engineering. Clean up there and you clean up twice.
Do we need a full lean program with boards and daily meetings?
No. Walking one order through costs an afternoon and gives you your top three wastes. Boards, daily meetings and fixed improvement rounds can always come later, once the first measures taste like more. Start small and concrete, otherwise lean itself becomes overprocessing.
Which waste should I tackle first?
The one with the most hours in your measurement. In practice, motion and defects often win: they can be sealed off technically and the gain is felt every single week. If you want to start there right away, you can try Thundercad free for 30 days.