Is three dimensional simulation worth what it costs to build?

Three dimensional simulation is the best demonstration in this category. A model of a warehouse where you can watch a forklift travel an aisle, a worker lift a box onto a machine, a conveyor back up behind a jam. Nobody in the room needs it explained.

It is also, by a distance, the part of a simulation project most likely to be quietly dropped once the work starts. That gap matters, because a feature that wins the demonstration and loses the project is not the same as a feature that is good.

The complaint

The pattern is consistent enough to state plainly: buyers who purchase on the strength of the three dimensional demonstration frequently do not build in three dimensions afterwards.

The reason they give is not that it is hard. It is that it is slow, and that the slowness buys nothing. Placing and orienting objects in space, modeling travel paths, dressing a scene so it reads as the real facility, is work with a real cost in weeks. And at the end of it, the throughput figure, the queue length and the bottleneck are the same numbers the two dimensional model produced.

The visual is not the answer. It is a presentation of the answer, and it is an expensive one.

Where it genuinely earns its cost

This is not an argument that three dimensional modeling is useless. There are cases where it is the right choice and no substitute will do.

Where space is the constraint. If the question is whether the aisle is wide enough, whether the racking blocks the turn, whether two vehicles can pass, then geometry is the problem and a model without geometry cannot answer it.

Where travel time dominates. A warehouse where pickers walk further than they work is a spatial problem wearing a process costume.

Where the audience will not believe a diagram. Some rooms will not accept a result they cannot see. That is a fact about the room rather than about the model, and it is a legitimate reason to spend the weeks.

If none of those apply, the three dimensional model is a rendering of a conclusion you already had.

What else persuades a room

The argument for three dimensions assumes the alternative is a spreadsheet. It usually is not.

A two dimensional animated run shows work arriving, queueing and clearing. People who have never seen a simulation understand a queue growing without being told what they are looking at.

A heatmap of when queues build is frequently more persuasive than watching them build, because it answers the next question. Watching a queue form tells a manager there is a queue. A grid showing it happens every Tuesday after two o’clock tells them what to do on Tuesday.

A chart ranking every step by the time and cost it adds settles arguments that an animation only starts. The step everyone blames is often fourth on the list, and that is the moment a project changes direction.

None of those take weeks to produce. They come out of the run.

How we score it

We score presentation as its own criterion, at ten percent. A product that cannot communicate a result has not finished its job.

What we have changed is what the criterion measures. It is not “does this have a three dimensional view”. It is how well a finished run persuades somebody who did not build the model. A three dimensional view is one route to that. An animated run, a heatmap, and a ranked chart are others, and they arrive with the result rather than costing weeks on top of it.

Under the old wording, a product with no three dimensional view scored close to a product with no animation at all, which was wrong: they are not comparable. Under the new wording the gap between the best and the worst presentation in this category is narrower, and the products that carry three dimensions still score highest, because the capability is real.

We have also stopped awarding full marks for it. The strongest three dimensional product in this comparison treats the visual as the product, and the cost of producing that visual is a cost the buyer carries.

What to ask a vendor

If you are being shown a three dimensional model in a sales meeting, two questions are worth asking before the end of it.

How long did this model take to build, and how much of that was the visual? The honest answer separates the engine from the rendering.

What does the same analysis look like without it? If the answer is the same numbers in less time, you have learned what the visual is for. If the answer is that the analysis is not possible without the geometry, you have learned that your problem is spatial, and you should buy accordingly.