Module 5: Industry examples

After this module you will be able to locate the qualifying work inside a typical software, manufacturing, engineering, life sciences or food project, state the uncertainty, and say where the qualifying work stops. Every example is a labelled hypothetical with a fictional company; rates and country rules are on the country pages linked at the end.

What this module covers

• Where the technical uncertainty usually sits in each of five sectors, and where it does not.
• Five labelled examples, one per sector, with the record that shows the work in each.
• Where the qualifying work ends, and how to draw that boundary in money.

One question, five sectors

The question is the same in every sector: was there something a competent professional in the field could not know or work out from available knowledge, and did you do systematic work to find out? What differs is where that work hides, what gets wrongly swept in, and which record shows it.

Software

Illustrative example. Larkspur Software Ltd, a hypothetical company, needs to cut search latency on its document platform from two seconds to under 200 milliseconds at 50 million documents. The lead engineer records in a dated design note that the four documented indexing strategies had been benchmarked and none met the target at that scale. Over four months the team builds a benchmark harness, implements two strategies with modifications, abandons one at six weeks when it fails on long queries, designs a hybrid index and reaches 180 milliseconds. That is the resolution of a technological uncertainty; the tickets, the benchmark runs including the failures and the design note are the evidence. The admin interface rebuild, the billing integration through the provider's documented API and the cloud migration involved no unknown a competent engineer could not resolve with known techniques, and are not claimed. The three further months of hardening and rollout after the benchmark was met are outside. Several of the schemes restrict or exclude software built for your own internal administration, so check the country page before assuming an internal tool qualifies.

Manufacturing

Illustrative example. Corran Castings Ltd, a hypothetical company, takes an order for a component in an aluminium alloy it has not cast before and finds in the first pours that porosity exceeds the specification. Nobody calls this research, but the engineers did not know what combination of pouring temperature, die temperature and degassing would hold porosity within specification for this alloy in their dies. Over three months they run a structured series of trial pours, varying one parameter at a time, scrapping 60 castings and recording each on a trial run sheet with the parameters, the result and the engineer's name. Pour 61 meets the specification and the parameter set is fixed; that is where the qualifying work ends. The production run, the routine quality inspection and the commissioning of a new furnace to the maker's specification are not claimed. The trial run sheets and scrap records are the evidence.

Engineering

Illustrative example. Fenwick Structures Ltd, a hypothetical consulting engineer, designs a long-span composite floor using a timber species for which no published shear connector data existed. The uncertainty is whether a connector could achieve the stiffness and capacity the span required with this timber. The engineers design two connector geometries, test 12 specimens to failure, discard one geometry when it splits the timber at 60% of the target load, and derive design values from the second. The connector programme is the qualifying work; the test reports, calculation iterations and dated design notes are the evidence. The standard beam design, drawings, code compliance checks and project management are not claimed. Engineering services carry one more question: who bore the technical and financial risk, the consultant or the client. Several schemes decide who can claim on that basis, so read the country page before assuming the claim is yours.

Life sciences

Illustrative example. Oriel Biosciences Ltd, a hypothetical company, needs a liquid formulation of a protein therapeutic that stays stable for six months at room temperature, where the known formulations required cold storage. Whether any excipient combination could achieve that for this protein was unknown. The team screens eight combinations against a written protocol, with stability samples pulled at two week intervals; six fail by week eight, two hold through six months, one is selected. The protocol, lab notebooks, batch records and analytical results are the evidence, and the work ends when the six month data are in. Routine release testing of later batches against the fixed specification is not claimed. Life sciences work usually passes the eligibility test easily; what decides these claims is where the work was done and who funded it, especially for trials run through contract organisations.

Food and drink

Illustrative example. Holloway Foods Ltd, a hypothetical company, runs two projects on the same cereal bar in one year. The first reformulates the bar to taste more like caramel. None of it involves scientific or technological uncertainty: the question was what people prefer, and several schemes exclude changes of taste, style or appearance expressly. It is not claimed. The second sets out to keep the bar from going stale for twelve months in a new packaging film without its current preservative, where the available data on water activity and oxygen transmission did not predict the outcome. The technologists run accelerated and real-time storage trials on five formulation and film combinations, with moisture, texture and microbial testing at intervals; three fail at four and seven months, one reaches twelve. The formulation log, storage trial schedule and test results are the evidence, and the work ends when the twelve month data confirm the result.

Illustrative example

Illustrative example. Drawing the boundary in money for Corran Castings Ltd. The whole alloy project cost 400,000. The trial pour programme was three engineers for three months, 100,000, plus 20,000 of metal and consumables scrapped in the 60 trial castings. The production ramp cost 200,000 and the new furnace 80,000. The candidate base for the claim is the 120,000 spent resolving the porosity uncertainty, subject to each scheme's rules on consumed materials and equipment; the 280,000 of production and plant is outside the boundary. The benefit at that country's rate applies to the 120,000.

Across the seven countries

One rule per country, in its own words, that moves the line in one of these sectors; the industries pages walk through every sector.

Where to go next

• Guide: R&D tax incentives for manufacturers
• Guide: qualifying versus non-qualifying activities
• Resource hub: eligibility
• Compare the seven schemes
• Technical Teams Program overview