Change the Design and Change the Demand

A pump has stopped. The building manager wants the water moving again before the morning shift. On a supplier’s screen, a replacement motor is available. It is cheaper than the old model and uses a different design. The manager could order it now.

The technician has a few questions. Will it fit the space? Can it turn the pump at the right speed? Will the existing controls work with it? Who will repair it, and where will the spare parts come from?

Nobody in the building came to work because they wanted a particular metal. They want water at the places where water is needed. Yet getting that service back depends on details that a list of elements cannot settle.

This is where substitution begins. A substitute has to do a job. Sometimes that means a new material in a familiar part. Sometimes it means a different part, with new controls around it. Sometimes the best change lies elsewhere: repair the leak, reduce the distance, choose a smaller machine. Each choice changes what must be bought from the mineral economy.

That gives us more room to act than a map of deposits suggests. We can alter demand as well as find supply. But the room is uneven. A building manager with a failed pump this morning has less freedom than a team designing a building for next year. Decisions made before trouble arrives can give the next person a choice that would otherwise be missing.

We should treat that choice as something worth building. The goal is a wider set of reliable ways to provide a good life. A new design earns its place when it expands those ways, at a cost people can understand and bear.

The part and the job

Consider the job of turning a shaft. A motor buyer can specify the work required without declaring that a particular magnet must do it. That small change in the question opens a much larger engineering discussion.

In February 2025, BMW described two motor types for its sixth-generation electric drive. One creates the rotor’s magnetic field through windings supplied with direct current. The other uses induction. Neither description depends on permanent magnets in the rotor. In August, BMW reported that series production of the new drives had begun at Steyr. These are the manufacturer’s accounts of its own design and production, rather than an independent comparison of all motor options.¹

The example establishes something concrete: the task can be approached through a different motor design. It leaves many buying questions open. A motor that suits one car may be wrong for a pump, and a good pump motor may be wrong for a car. Even within one use, changing the design asks us to look beyond the replaced part.

Return to the building. Suppose an alternative motor requires more room, while another fits but needs different controls. The first choice may involve rebuilding the mounting. The second may involve an electrician and new software settings. A low purchase price could survive those costs, or disappear under them. The answer comes from the installation, not from the word alternative.

There may also be a trade between the cost of making the machine and the cost of running it. If one option costs less to build but uses more electricity at the building’s usual workload, someone has to count the years of operation. If the opposite is true, a larger up-front payment could lower later bills. These are possible comparisons to investigate, not fixed rankings of motor types.

Now suppose the supplier offers a very efficient design that only its own staff can service. The buyer has reduced one kind of dependence and accepted another. That may be a fair bargain. It should at least be visible. A machine can be free of a named mineral while remaining costly to leave.

This is why substitution should be judged at the level where the service is delivered. Material content still matters. So do access to repair, the skills needed to operate the equipment, and the terms on which a second supplier can step in. Put those questions in the order form early, while competitors can still answer them.

A buyer who waits for a shortage before asking has handed time to the supplier. A buyer who has already tested another way to do the job can discuss price with a real option behind the discussion.

What a new battery name tells us

Battery names can make substitution sound like a change of ingredient in a recipe. Take out one thing, add another, and keep the rest of the meal.

The names do reveal something. Lithium iron phosphate, usually shortened to LFP, uses a cathode material without nickel or cobalt. The cathode is one of the battery’s two working electrodes; its chemistry supplies the names used here. NMC names a cathode containing nickel, manganese and cobalt. Both belong to the lithium-ion family. A working battery system needs other materials too, along with controls and heat management. Its chemistry affects performance, while the full design and the way it is used also matter.²

So removing cobalt from that cathode does not remove lithium from the battery. Nor does the new name tell a buyer where the remaining inputs are processed. The material list has changed. There is still a supply chain to examine.

A 2025 teardown study makes the comparison more specific. Jonas Gorsch and colleagues examined a BYD Blade LFP cell and a Tesla 4680 NMC cell. The reported energy per kilogram of cell was about 160 watt-hours for BYD and 241 for Tesla. Both also used graphite. These were two particular cell designs obtained in 2023 and from a 2022 vehicle. The result neither ranks every current product nor isolates chemistry from the other design choices. It does show why the complete cell deserves examination.³

Imagine two complete storage systems offered to a small factory. Each can deliver the electricity the owner needs, under the stated conditions. One occupies more space. The other costs more. The factory has an unused yard, so the larger footprint may be easy to accept. Put the same choice inside a tightly packed vehicle and space carries a different price.

Even in the yard, the footprint is not free. It may compete with a loading area or leave less room for future work. The owner must ask about safe installation, maintenance and the cost of replacing parts. The comparison should use the same service: how much energy reaches the equipment, when it is needed, over the agreed life. A cheap unit that cannot meet that duty is not the cheap way to meet that duty.

This sounds obvious until a mineral slogan enters the room. Then the name of an avoided element can take over the whole discussion. A material change deserves credit for the problem it solves. It should not receive credit for every problem in the system.

Sodium-ion research makes the point from another direction. In a 2025 study, Adrian Yao and colleagues modelled sodium-ion costs against lithium-ion alternatives under changing technical and market assumptions. A sodium cost advantage was conditional. Better energy density helped its prospects; mineral prices and the chosen chemistry also changed the result. The authors cautioned against reading the model as a prediction of an exact crossover year.⁴

That is a stronger basis for hope than an automatic victory. It tells researchers and buyers where work still has to be done. A promising material needs a promising complete cell. A promising cell needs a repeatable production process. A repeatable process needs customers who can use what it makes.

There is another way to read the uncertainty. Suppose a buyer has qualified two designs, both adequate for the job. One is cheaper under ordinary conditions. The other uses a different set of inputs. Keeping the second option available could be worthwhile even if it does not win every normal purchase. Its value includes the choice it preserves when conditions change.

How much should that choice cost? There is no single answer. The buyer can compare the cost of keeping an alternative ready with the loss caused by an interruption. A hospital’s backup equipment and a household’s spare gadget will justify different spending. We lose that judgment if we use either cheapest today or secure at any price as a complete buying rule.

The trip we meant to make

We can also change the service around a product.

Picture a business choosing vehicles for a delivery route. The team starts with the heaviest load and longest day it can imagine. It selects a vehicle able to meet both, even though they rarely occur together. Every ordinary day then carries the cost of that choice.

Perhaps this is sensible. Customers may need a promise that holds without warning. A missed delivery could cost more than carrying spare capacity. But perhaps the business could divide its work differently. It could use a smaller vehicle for the regular route and hire extra capacity on the unusual days. It could change the loading schedule. It could place a collection point closer to customers who want one.

Each proposal must survive the actual job. Hiring extra capacity is useful only if it will be available. A collection point can save a driver time while shifting a longer journey onto the customer. A smaller vehicle could require more trips, wiping out the expected gain. The team should count the whole service, including the burden moved onto other people.

This is a demanding form of material efficiency. It asks for less material per successful delivery, rather than less material in a vehicle considered alone. The difference protects us from a tidy saving that makes life worse elsewhere.

It also brings people with little say in product design into the decision. The driver knows which streets cause delays. The customer knows whether a new collection point is reachable. The person who repairs the vehicle knows which part keeps failing. Their knowledge can reveal a better design before anyone searches for a new mineral.

Now take a shared tool store in an apartment building. If residents can borrow a drill when they need one, some may choose to own fewer drills. But the service must be real. A locked cupboard with no reliable booking system will not create much confidence. Someone has to handle repairs and missing parts. Residents who work late need a way to collect the tool outside office hours.

The material saving depends on what people actually stop buying, how long the shared tools last, and what work the store adds. Merely placing tools in a cupboard proves none of that. A well-run service could still be a gain worth pursuing. It would make occasional access affordable while giving maintenance work a regular home.

Such changes should widen access, not turn material restraint into a lecture directed at people who already have too little. A family obtaining its first dependable refrigerator has a different need from a household replacing a working one for a fashionable finish. Counting objects without asking what they do would treat those choices as if they were the same.

The practical ambition is more service from each object, with more people able to use the service. That can support growth in well-being while easing a particular material demand. It also explains why lower material use per unit does not guarantee lower total use.

Suppose a product is redesigned to use less metal. If sales stay the same, that change reduces the metal needed for those sales. If the lower cost makes the product affordable to many new buyers, total sales may rise enough to offset the saving. Some of the benefit has returned as greater use. This possible response is often called rebound.

Whether that greater use is welcome depends on what it provides. More access to a needed service can be a social gain even when total material demand grows. More needless replacement is harder to defend. Efficiency should be assessed with both facts in view: the material used for each unit and the number of units made. We can celebrate a better design while remaining honest about the scale of the task.

Make room for a second answer

A substitute on paper offers limited bargaining power. Someone must show that it works in the product, and someone must be able to supply it when asked.

For the pump buyer, that means testing the relevant duty rather than admiring a demonstration. How does the proposed replacement behave when demand changes? Can the maintenance team find a fault? What happens if a control fails? The checks should match the risk. A trial on a non-essential machine can answer questions that would be reckless to discover on the building’s only water supply.

This work costs money before a shortage makes its value obvious. If the buyer never budgets for testing, its purchasing policy may praise competition while leaving only one usable supplier. The entry barrier is partly in the buyer’s own hands.

A fair qualification process can widen the field. Write down what a part must do. Explain the test. Give a prospective supplier a way to correct a failure and try again. If the requirement protects safety or reliability, keep it. If it merely copies the shape of the incumbent’s product without serving a need, ask why it is there.

The same care applies to public support for a new technology. A promise to replace a scarce input is a reason to investigate, not an unlimited claim on public money. A trial can be tied to a clear performance question. The next payment can depend on the answer. Money can support a shared testing facility rather than require every small firm to buy the same expensive equipment.

That spreads the chance to improve a design. It can also leave a lasting skill after one candidate fails. A laboratory able to measure product quality still has a job when a particular material loses its market. Workers who learn testing, fault finding and process control gain knowledge that may travel across designs.

We should plan for that travel. A factory built around one material cannot move its workers into a new process by announcing that the future has arrived. The workers need training and time; their households still need income while the equipment changes. If society benefits from reducing a supply risk, there is a case for sharing the cost of that transition instead of placing it all on the people whose old work is being displaced.

The producer of the original mineral faces a related choice. If a customer changes its design, demand may shift. A development plan that spends every expected future royalty in advance leaves little room to respond. Investing some present income in skills and businesses that serve more than one industry can preserve choices. The value of a resource boom then extends beyond the material that started it.

At the building, the technician finally recommends a replacement. It will do the required work, the controls can be adapted, and another firm can supply the service parts. The price includes the installation. The manager can now compare complete offers.

The water matters most to the people arriving for work. Behind it sits a less visible gain: the building no longer needs exactly one answer to keep functioning. Good substitution makes that gain possible. It changes a demand that once seemed fixed into a choice that people can test, improve and share.

Sources and limits