Quick answer: Difficulty is not the test. A core R&D activity requires an outcome that could not be known or determined in advance on the basis of current knowledge, information or experience, and could only be determined by a systematic progression of work. Bespoke engineering on difficult sites — reactive soils, deep basements beside existing structures, wind- or seismic-critical forms, heritage retention — may still involve a competent professional applying published codes and established methods, in which case the outcome may be determinable in advance even though the work is demanding. A potential core activity arises only where the relevant behaviour could not be determined from the field's current knowledge and required systematic experimental work. You self-assess.
7 August 2026 — this article describes the current rules. The 2026–27 Federal Budget announced proposed R&DTI reforms for income years starting on or after 1 July 2028. Until any amendments take effect, the R&DTI continues to operate under the current rules.
A structural or geotechnical project can absorb months of senior time, run through more design iterations than anyone budgeted for, and still fall outside the R&D Tax Incentive (R&DTI) entirely. The program does not respond to difficulty, cost, iteration count or professional risk. It asks whether a competent professional could have known or determined the outcome in advance from existing knowledge, information or experience that was publicly available or reasonably accessible — not simply whether the answer was new to the project team.
This article is confined to structural and geotechnical engineering judgement. Two adjacent questions have their own articles in our Insights: the general separation of business risk from technical uncertainty, and how National Construction Code compliance testing sits against developing the building system you are testing.
Where the Unknown Outcome Has to Sit
Core R&D activities are experimental activities whose outcome cannot be known or determined in advance on the basis of current knowledge, information or experience, but can only be determined by applying a systematic progression of work that is based on principles of established science and proceeds from hypothesis to experiment, observation and evaluation, and leads to logical conclusions — and which are conducted for the purpose of generating new knowledge (business.gov.au). The established-science limb does independent work on building sites: a build-it-and-see loop through successive configurations can be disciplined, expensive and well documented and still fail the test, because trying arrangements until one holds is not a progression grounded in the mechanics, soil behaviour or materials science that would let a hypothesis be framed and a result interpreted. Three practical questions do most of the work when that definition meets an engineering project:
Which field is the relevant one? The knowledge standard is the field's, not the practice's, so the field has to be identified before anything can be said about what it knows. In construction work the boundary is rarely obvious: a modular connection problem may sit in structural steel connection behaviour, in cold-formed steel, in cyclic fatigue, or in soil-structure interaction, and the published state of the art is different in each. Where the field is drawn narrowly enough — "modular connections in South Australian reactive clay" — almost anything looks novel. Drawn at the level a competent professional would recognise, most of it does not. The relevant field should be framed broadly enough to reflect the knowledge and experience reasonably available to a competent professional addressing the technical question, rather than being defined so narrowly around the particular project that ordinary engineering appears novel.
Was the model operating inside its validated envelope? Established analysis methods have documented ranges of validity: the geometries, materials, load types, strain ranges and boundary conditions the method has been calibrated and verified against. Where an established analysis method is being applied within conditions for which its behaviour is well understood, that may support a conclusion that a competent professional could determine the relevant outcome in advance. Conversely, operating outside a method's established or validated range may indicate a technical uncertainty, but does not by itself establish core R&D; the activity must still satisfy the complete statutory requirements. Writing down the envelope, and where the project sits relative to it, is the step most engineering files skip.
Is this calibration or experimental validation? The two produce nearly identical documents and mean opposite things. Calibration fits the parameters of an accepted model to site or material data so the model can be applied — the model's validity is assumed, and the data resolves an input. Experimental validation tests whether a relationship holds at all, with a result that could show the model does not describe the behaviour. Instrumented monitoring next to an excavation is usually the first: the prediction exists, and the readings confirm it. A monitored trial designed so a stated hypothesis about ground response could be disproved is closer to the second.
Where the Line Usually Falls in Structural and Geotechnical Work
Eligibility questions tend to concentrate on whether the achievable physical behaviour was unknown to the field, rather than on whether the design solution was hard to arrive at.
Usually not a core R&D activity
May contain a core R&D activity
Founding on reactive or expansive clay using site classification and design procedures in the applicable Australian Standard, however severe the site
A ground-improvement or foundation treatment whose achievable behaviour in that ground was not predictable from published data or established method, resolved by an instrumented trial with a stated hypothesis
Deep basement, shoring and retention adjacent to existing structures, with movement predicted by established geotechnical models and monitored against them
A retention or support system whose interaction behaviour falls outside the validated range of any available model or precedent, where the achievable performance could not be determined in advance
Wind-critical or seismic-critical forms designed and, where required, wind-tunnel tested under the applicable loading standards
A hybrid or composite structural system with no applicable design code, where load path, ductility or connection behaviour cannot be derived by extension from existing codes
Heritage and façade retention, temporary works and staged demolition designed by experienced judgement and established methods
A new material or component whose structural properties under the relevant actions are not established in the literature or in certification data
Remediation of contaminated ground to a regulator's requirements
A performance-based approach where no code path or accepted verification method exists and the achievable behaviour was unknown
Reproducing or reverse-engineering an existing proprietary system from inspection, plans or specifications
Connection and diaphragm behaviour in a volumetric or modular system where transport, lifting and assembly states are not covered by any applicable code
The absence of a directly applicable design code does not by itself put an activity in the right column. Engineers routinely and correctly derive answers for uncodified arrangements by extension from first principles and adjacent codes; what matters is whether that derivation could reliably determine the outcome, or could only generate a hypothesis that had to be tested. A performance solution is likewise not automatically an experiment: choosing a performance-based pathway over a prescriptive one is a compliance choice, and where an accepted verification method exists and was applied, the activity looks like compliance with a standard. Our companion article on compliance testing versus developing the system being tested works through that boundary; see also R&D for property and construction.
The Exclusions That Catch This Work
Section 355-25(2) of the Income Tax Assessment Act 1997 lists categories that cannot be core R&D activities at all. Two recur in this sector, as set out on the department's excluded R&D activities page: activities associated with complying with statutory requirements or standards, which may include site classification, certification or conformance testing where those activities are undertaken to satisfy a statutory requirement or standard and reproducing a commercial product or process by physical examination or from plans, specifications or publicly available information, which covers working out how a competitor's proprietary connection or panel system behaves in order to build your own.
The practical difficulty is that the same rig, the same specimens and often the same engineer can produce both a conformance test and an experiment in the same week. Where a development program exists, the certification testing that follows it has a separate purpose: one asks whether a relationship holds, the other demonstrates to a certifier that a settled arrangement meets a nominated requirement. Booked to one job number, that distinction is not visible in the record afterwards.
Supporting activities and dominant purpose: Activities that are not core may qualify as supporting R&D activities under s 355-30 where they are directly related to core R&D activities — and, where the activity is of an excluded kind, produces goods or services, or is directly related to producing goods or services, only where it was undertaken for the dominant purpose of supporting the core activity (business.gov.au). In construction the production limbs are unavoidable: the project produces a building, so most site work either produces goods or is directly related to producing them, including installation and delivery work that is not itself production. If there is no core activity, there is nothing for supporting activities to attach to.
A Worked Example
Illustrative and hypothetical only — not a ruling, not a description of any client, and not a statement that any of it would be eligible.
An Adelaide modular builder is developing a three-storey volumetric residential system. Modules are steel-framed, 3.6 m wide and 12.0 m long, stacked three high on a site assumed, for this illustration, to have a characteristic surface movement of 75 mm, at the upper end of the H2 highly reactive range.
Designing the raft and articulation by applying an established standard may already fail the core R&D test where the outcome can be determined in advance. Separately, activities undertaken to comply with a statutory requirement or standard may also fall within the specific s 355-25(2) exclusion. The same is true of the shoring design for the podium excavation beside an existing building, where established geotechnical models predict movement and the monitoring confirms the prediction rather than resolving an unknown.
Testing the inter-module connection: The inter-module connection is different. No applicable design code covers it, and the question the team cannot answer from the codes or the literature is whether a connection can transfer hogging moment across the module joint while remaining serviceable through repeated seasonal ground movement. The illustrative load case is a differential support settlement of 25 mm across the 12.0 m module length, cycled seasonally. Before any specimen is built the team records the hypothesis and a pass criterion: the connection must retain at least 90% of its initial rotational stiffness after 200 cycles of ±0.004 radian imposed rotation, with no visible weld-toe cracking and residual bolt slip below 1.5 mm. Six full-scale specimens are fabricated, two per configuration.
Configuration 1: A bolted end-plate with standard clearance holes, loses stiffness from cycle 40 and reaches 55% by cycle 90, with 4 mm of residual slip. That result rules out connections relying on clearance-hole bolting to hold rotational stiffness under repeated cycling, and it also rules out the assumption behind it — that slip could be neglected at this rotation.
Configuration 2: A fully welded splice, holds stiffness but develops a weld-toe crack at cycle 120: rigidity moves the demand into the weld rather than removing it.
Configuration 3: Responds to both failures by combining preloaded friction-grip bolts with a replaceable yielding shear link intended to take the cyclic demand away from the weld. One specimen completes 200 cycles at 92% stiffness; the second reaches 88%, below the pre-set criterion. On those facts the question is unresolved and further testing would follow.
The activity requiring assessment is much narrower than the project: the experimental development and testing of the connection. Eligibility still depends on the complete facts and statutory tests. Detailing, shop drawings and module production are production work; some directly related activity may warrant assessment as supporting, but only where the dominant purpose test in s 355-30 is satisfied on the facts.
What the Record Shows, and What It Does Not
Engineering practices generate large volumes of documentation and very little of it records an inquiry, because it is written to discharge a duty of care. Calculation packages, RFIs and design certificates describe what was designed, not what was unknown. AusIndustry sets out its expectations for records to show eligibility, and registration is self-assessed.
In projects we assess, the records that most clearly expose the boundary are these:
• A contemporaneous search record showing what sources were considered and how the asserted knowledge gap was framed at the time;
• A contemporaneous test plan recording the hypothesis, evaluation criteria and experimental method to help establish the scope and timing of the experimental activity;
• Superseded configurations and the raw instrumentation traces behind them, versioned rather than overwritten, as surviving evidence of what was ruled out;
• A contemporaneous note recording the conclusion reached or the decision to discontinue a line of inquiry to help evidence when that experimental phase concluded.
A practical record-keeping risk is merged evidence: connection development, certification testing and project delivery all booked to one job number, so no method exists afterwards for separating engineering hours spent resolving the unknown from hours spent demonstrating conformance or delivering the building. How that separation is treated in a claim, and how the expenditure is characterised, is a matter for the company and its registered tax agent.
Where an RSP Fits
AusIndustry describes Research Service Providers as scientific or technical service providers you can engage to conduct R&D activities on your behalf, registered in specific research fields (business.gov.au). For a builder or engineering practice, an RSP may assist with experimental design, technical R&D work and contemporaneous supporting records within the research fields for which it is registered. Earlier involvement can help document the technical question, hypothesis and experimental approach as the work progresses.
Smaller test programs run into the entitlement floor. R&D expenditure for the income year must generally be at least $20,000. However, qualifying expenditure incurred to a non-associate RSP may still be taken into account in determining R&D tax offset entitlement where total notional deductions are below the usual $20,000 threshold. The substituted base in s 355-100(2) of the ITAA 1997 is generally limited to qualifying expenditure incurred to a non-associate RSP for services in a field for which it is registered, together with eligible monetary contributions under the CRC program (ATO). See claiming R&D under $20,000. Using an RSP does not guarantee eligibility — you still self-assess. Offset rates, the refundable and non-refundable tiers and the intensity premium are covered in refundable vs non-refundable offset.
Talk to Ignition Research before you register a difficult project as R&D. As a Registered Research Service Provider at Lot Fourteen in Adelaide, we work with builders, developers and engineering practices across South Australia to identify the relevant field, test whether the knowledge gap sits in the state of the art, and design the experiment so the record exists in advance. We supply research capability, not tax advice: we are not a registered tax agent, your company self-assesses and remains responsible for its own claim, and advice and lodgement stay with your tax adviser. Get in touch.
Frequently Asked Questions
Q: Is difficult engineering on a complex site eligible for the R&D Tax Incentive?
A: Difficulty is not the test. A core R&D activity requires an outcome that could not be known or determined in advance on the basis of current knowledge, information or experience. Where a competent professional could have determined the outcome by applying published codes, standards or established analysis methods, the activity is generally not core R&D — however demanding, expensive or iterative it was. You self-assess.
Q: Does designing for reactive soils or a deep basement count as R&D?
A: Not automatically. Site classification and foundation design on reactive clay follow published Australian Standards, and retention and movement prediction next to existing structures use established geotechnical models. Both are competent professionals applying current knowledge. Work may warrant assessment as core R&D only where the achievable ground or system behaviour was not predictable from published data or accepted method, and was resolved by systematic experiment.
Q: When is a new structural system eligible R&D?
A: It may qualify, subject to self-assessment against all the statutory requirements and exclusions, where there is no applicable design code and the behaviour — load path, ductility, connection response, long-term performance — could not be derived by extension from existing codes or literature, so it could only be determined by a systematic progression of work from a stated hypothesis, conducted for the purpose of generating new knowledge. Novel geometry alone, designed by established methods, generally will not.
Q: Does it matter that the problem was new to our engineers but not to the profession?
A: Yes. The relevant question is not simply what your own team knew. You need to assess whether a competent professional could have known or determined the outcome in advance using knowledge, information or experience that was publicly available or reasonably accessible at the time, including relevant sources available internationally. A problem that is new only to your team will not satisfy the unknown-outcome requirement merely for that reason.
Sources & Further Reading
ATO — Eligibility for the R&D tax incentive — the $20,000 floor and the substituted RSP/CRC base in s 355-100(2)
legislation.gov.au — Income Tax Assessment Act 1997 — Div 355, incl. ss 355-25, 355-30 and 355-100
Related: R&D for property and construction · what does not qualify · what an RSP is · refundable vs non-refundable offset · claiming R&D under $20,000 · more Insights
This article is general information from a Registered Research Service Provider about the R&D Tax Incentive. It is not tax, legal or financial advice; eligibility depends on your circumstances and you should self-assess and seek your own advice.
Thinking about a project like this?
If you're weighing up an AI, software or technical improvement project and can't tell yet whether it's implementation or research, start with a quick read on where it sits.