Feedback design, embedded controls and industrial automation: identify experimentation, not routine commissioning.
Estimate your credit →Control systems development may qualify when engineers resolve technical uncertainty in a new or improved process, product or software component through experimentation. Comparing control strategies, system models or real-time architectures can be relevant when all research requirements are met. Standard programming, installation and parameter adjustments using known methods generally are not enough.
Only documented qualified research expenses count. Eligible costs may include research wages, supplies used in research, qualifying computer rental and generally 65% of qualifying contract research. Production hosting, depreciable equipment and routine operations are not automatically eligible.
This guide focuses on deterministic feedback, PLCs and industrial automation. For learned perception, AI models and autonomous robotics, use the AI + Robotics guide. Explore AI + Robotics
These activities may qualify only when the project meets the four-part test and applicable exclusions. A job title, new product or use of AI does not establish eligibility.
| Business component | Example | Technical uncertainty |
|---|---|---|
| Feedback controller design | Compare PID and model-predictive approaches | Which design remains stable under varying loads and delays? |
| PLC logic architecture | Prototype synchronization logic for interacting machines | Can the architecture satisfy cycle-time and fault-recovery requirements? |
| Embedded real-time controls | Compare scheduling and communication approaches | Can timing deadlines be met despite processor and network limits? |
| System identification | Develop and test a model of a thermal process | Which model captures nonlinear behavior accurately enough for control? |
| Motion and servo control | Evaluate compensation methods on a prototype axis | Can the system maintain precision despite vibration and mechanical variation? |
| Fault detection and recovery | Compare sensor diagnostics and recovery logic | Which design detects faults without unacceptable false trips? |
| Hardware-in-the-loop testing | Simulate process disturbances against controller prototypes | Will the proposed control design remain stable at operating limits? |
| Industrial communication | Prototype synchronization across heterogeneous controllers | Can latency and packet loss be handled without losing process stability? |
Find your own job titles below. Each row shows the wage category a role usually falls under and the kinds of work that typically do and do not count. Titles alone never make wages eligible; the activity must meet the four-part test.
6 of 6 roles shown
| Role | What they do | Wage category | Typically qualifies | Typically does not |
|---|---|---|---|---|
| Designs how equipment responds to changing conditions. | Direct research | Comparing feedback strategies to resolve uncertainty, Experimental stability and response testing | Routine controller tuning, Applying a known design unchanged | |
Typically qualifies
Typically does not
|
||||
| Builds logic that coordinates machines and processes. | Direct research | Testing uncertain synchronization architectures, Experimental fault-recovery logic | Standard ladder-logic configuration, Routine commissioning | |
Typically qualifies
Typically does not
|
||||
| Builds software that runs inside the controller. | Direct research | Evaluating uncertain scheduling designs, Prototype timing and communication experiments | Routine firmware updates, Porting an established configuration | |
Typically qualifies
Typically does not
|
||||
| Models process behavior before hardware trials. | Direct research | Comparing uncertain plant models, Experimental controller simulation | Routine reporting, Reusing a validated model without experimentation | |
Typically qualifies
Typically does not
|
||||
| Sets up sensors and records experimental results. | Direct support | Instrumenting qualified pilot experiments, Collecting test data directly supporting research | Scheduled calibration, Production maintenance | |
Typically qualifies
Typically does not
|
||||
| Directs the team's day-to-day technical research. | Direct supervision | First-line supervision of qualifying control experiments, Reviewing technical alternatives with researchers | Budget administration, Sales meetings and general management | |
Typically qualifies
Typically does not
|
||||
Wage categories follow IRC Section 41. Only the portion of time spent on qualified research counts, and the percentage must be documented.
Engineers do not know which control strategy can maintain temperature as feedstock and load change. They compare process models and feedback designs in simulation, then run instrumented pilot trials. The records connect each alternative, test result and design change to the initial uncertainty.
Review documented research time and eligible test supplies separately from standard PLC installation, routine tuning, production runs and operator training. Customer contract terms and research location also matter.
Illustrative project, not a credit estimate or eligibility determination. Document the uncertainty, alternatives, results and eligible expenses before calculating a credit.
The IRS requires qualifying research to satisfy four tests. Here's how they typically map for this industry:
Improve the function, performance, reliability or quality of a controller, production process or embedded system.
Rely on engineering, physics or computer science to model process behavior and develop control logic.
Identify uncertainty about stability, timing, sensing or appropriate control architecture before development begins.
Compare alternative models, algorithms or architectures through simulation, hardware-in-the-loop tests and instrumented trials.
Take these into a 15-minute conversation with your team.
The last question matters most. Customer-funded work and work performed outside the US are generally excluded. Funded work is excluded to the extent your business does not bear the financial risk or does not retain substantial rights in the results.
No. Standard programming or configuring known logic generally is not qualified research. The project needs technical uncertainty and a process of experimentation, along with the other requirements.
Routine commissioning and tuning using established procedures generally do not qualify. Distinct experiments evaluating uncertain control designs may warrant review, with their time and costs separated from ordinary installation work.
It depends on the contract and project facts, including who bears economic risk and retains substantial rights. Customer-funded work requires a funded-research review rather than assuming that all custom engineering qualifies.
Simulation and hardware-in-the-loop trials can support a process of experimentation when they evaluate technical alternatives. A failed test does not automatically disqualify the work, but records must show what uncertainty was investigated and what was learned.
Documented qualifying wages, eligible nondepreciable research supplies and generally 65 percent of qualifying US contract research payments may count. Depreciable controllers and production machinery are not qualified supplies, and non-US research generally is excluded.
Keep requirements, control diagrams, model comparisons, code revisions, test plans and trial results tied to each business component. Support time allocations and costs, and distinguish experimental development from routine operations. Review the four-part test and expense rules
Most companies underestimate what they can claim. Our study covers every qualifying role, every component, and every year still open for amendment.
Get your free estimate →