Buying exhibits is not like buying equipment. Equipment has a specification, a datasheet and a competitor selling the same thing. An exhibit has a drawing, a behaviour, and a price that depends on decisions you have not made yet. This guide is the sequence we walk through with a client before a quotation means anything.
The most common way a STEM floor goes wrong is that it is planned around a topic — space, robots, water — and the age of the visitor is decided afterwards. The topic is the easy part. The age decides the mechanism, the height of the panel, the reading level, the number of users per unit and the price.
Ask three questions and write the answers down:
Those three answers decide more of the specification than any design review. A unit built for a seven-year-old and one built for a fourteen-year-old can share a topic and share nothing else: different mechanism, different enclosure height, different graphics, often a factor of two in price.
Area is the worst unit for planning an exhibit list, because a hall can be large and still only hold four themes, and four themes is what a visitor can absorb in ninety minutes. Our own reference project, a science centre in East Java, runs about 18,000 m² of gross floor area with nine subject themes and just over two hundred interactive units. That is roughly twenty units per theme, and it is a more useful ratio than any per-square-metre figure.
Practical ranges we quote:
If a plan has fifty units spread across twelve themes, it has four units per theme, and four units is not a theme — it is a display. Fewer themes done properly beats more themes covered thinly, and it is also cheaper to maintain.
Two quotes for "the same" exhibit can differ by a factor of three without either supplier being dishonest. What changes is the structure underneath. When you compare, break the number into these five parts:
| Part | What it covers | Why it varies |
|---|---|---|
| Mechanism | Frame, moving parts, drive, bearings | Duty cycle: a school unit and a holiday-centre unit are different machines |
| Control | Electronics, sensors, software, wiring | Mechanical units have none; interactive digital units can double the cost |
| Finish | Paint, graphics, acrylic, edge work | Print quality and panel replacement strategy |
| Packaging | Crates, foam, desiccant, loading plan | Sea freight needs export crating; air freight does not |
| Installation | Site labour, commissioning, training | On-site team versus remote commissioning with a local crew |
A quotation that is one number with no breakdown is not comparable to anything. Ask for the five lines, then ask which lines would change if the unit ran six days a week instead of seven.
Every unit on a public floor has a wear item: a ball, a panel, a bearing, a lamp, a set of building members. The question is not whether it will fail but whether the museum can replace it without waiting for a shipment and an import clearance.
Our standard practice is to ship a spare kit with the order — consumables plus one of each critical module — and to put a part number on the part itself. Museums that skip this line save two or three per cent on the purchase and spend far more in the first year of operation. Put spares in the budget at the start, where it is a rounding error, not in year two, where it is a purchase order and a delay.
Exhibit production is not the long part. The long parts are approvals, shipping and site readiness. A realistic sequence for an export project:
Work backwards from the opening date and you will usually find that the design approval is the item under pressure, not the factory. A factory that can start earlier cannot help if the drawings are still being revised.
The cheapest insurance on a project is a written list of what "working" means, agreed before shipping. Ours typically covers:
None of this is exotic. It is the difference between a delivery and a handover.
Buying a centrepiece first. The signature piece is usually specified before the age band and the layout, then everything else has to work around it. Specify it third or fourth, once you know what the floor is for.
Treating a display as an exhibit. A model behind glass teaches nothing and costs almost as much as a unit visitors can touch. If the budget is tight, fewer interactive units beat more static ones.
Ignoring the queue. Two visitors at a panel and two waiting behind them need more floor than the machine itself. A layout that fits the units but not the people will feel crowded even when every exhibit is working.
Underestimating the cleaning and reset routine. Water units, ball units and soap units need a daily routine. If the routine is not designed and staffed, the units will be switched off within a month.
Comparing prices without comparing duty cycles. The cheapest quote is often a unit built for occasional use, which is the wrong machine for a school that opens five days a week.
The most useful first message contains four things: the age band, the weekly opening hours, the floor area with a rough plan, and the budget range — even a wide one. With those, a manufacturer can tell you honestly whether your plan is achievable, and which themes to cut. Without them, every reply is a brochure.
We answer with a unit list and a budget range, marked with what we make and what we would change. If a layout cannot be delivered within the budget, we say so at that stage rather than after the drawings.
Written by our engineering and sales team from live project work, not from a survey. Figures describe our own projects and are indicative; they are not a quotation. Sizes, ages and specifications refer to standard models and change with custom work.
Standard models with fixed dimensions, control packages and spare-part lists. Custom sizes on request; single-unit orders accepted.

Bernoulli Visualization — an interactive science exhibit.

A platform shakes at recorded intensities while a seismograph traces the movement.

Build a span from interchangeable parts and load it until it fails.
Three of these are written for the same decision — what to buy, what it costs and what survives.
Attach a theme outline, a floor plan, or nothing at all — a photo of the space is enough to start. Quotations are issued against a real exhibit list, not per-square-metre guesses.
Prefer to see the floor before you commit? Factory visits and third-party inspections are welcome.