SportsTech · Case study

Research allowance for computer vision: real-time AI, funded.

Practical example based on real projects · anonymised, figures rounded · Last updated: 24 September 2026

A sportstech startup funded real-time video analysis on its own edge hardware, on an eligible base of around €1.3 million, entirely in-house, over a long horizon.

€380,000

research allowance granted

€1.3M

eligible base

Up to 35%

SME funding rate

Get a free funding assessment Read the story

At a glance

The mandate in figures

IndustrySportsTech / visionComputer vision on edge hardware

Research allowance granted€380,000In-house work only

Eligible base€1.3M

ScenarioCertified, long horizon

Projects1

SME funding rateUp to 35%

About the company

The company, and the R&D underneath

The company develops real-time video analysis for amateur sport that runs on its own modular hardware setup. The appeal and the difficulty lie in the combination: reliably recognising movements, game situations and relevant events in real time, not in a data centre, but directly on compact, low-cost hardware at the sideline.

This combination of sophisticated image processing and tight hardware budgets is not a solved standard problem. It demands its own development decisions with an open outcome, the funding core.

The challenge

Where the funding really sat

Computer-vision applications have two typical stumbling blocks. First, the delineation from the state of the art. Image processing is a mature field; much is available. Eligible is only what goes beyond it, here reliable real-time analysis under the tight conditions of proprietary edge hardware. This specific difficulty must be clearly worked out.

Second, the clean capture of in-house work over a long funding horizon. If the funding hangs solely on personnel costs and the project runs over many years, the eligible personnel work must be captured and assigned year by year.

Novelty Technical risk / uncertainty Systematic approach

Our approach

How we built the case

We built the project around the actual uncertainty: reliable real-time image recognition within the compute and energy limits of compact hardware. We researched the state of the art at the time of the project start and named the established methods against which the in-house development sets itself apart. The solution path was presented as a causal chain with metrics, recognition quality, latency, resource budget.

We captured the in-house work cleanly across the full, multi-year term and assigned it to the fiscal years. The project was cleanly certified.

How the case moved

↻

Scoped

Real-time under tight budgets

↻

Documented

In-house work, year by year

✓

Certified

Long funding horizon

What made up the eligible base

In-house personnel work 100%

In-house personnel work

Eligible base around €1.3 million, entirely in-house work over fiscal years 2022 to 2025. Figure shown is illustrative.

The result

A €380,000 research allowance.

The eligible base is around €1.3 million across fiscal years 2022 to 2025, entirely from in-house work. At 25% on the costs up to March 2024 and the 35% SME rate after that, this yields around €380,000 of research allowance. Thanks to the long funding horizon, the research allowance accompanies development over several years.

Eligible base €1.3M · SME rate Up to 35%

Certified, long horizonThe outcome of the mandate.

Eligible base €1.3MRecognised cost the allowance is calculated on.

SME rate Up to 35%Applied to the eligible base.

Key takeaways

What other companies can learn

Computer vision on proprietary hardware is a good funding case, with the right sharpening. Three points are decisive:

01

Name the specific difficulty

Not "we do image processing", but the concrete uncertainty, here real time under tight hardware budgets, carries the application.

02

State of the art at project start

In a mature field, the temporal anchoring decides novelty.

03

Capture in-house work year by year

With long terms and in-house work only, the clean, yearly assignment of personnel costs is the central lever.

More on Research Allowance: The guide Research Allowance: requirements, rates and application explains the programme step by step. How we support your project is covered under Research Allowance consulting.

“

Real-time image processing on small, proprietary hardware is a doubly demanding field: sophisticated algorithms meet tight compute and energy budgets.

BeFunded On the research allowance

FAQ

Research Allowance: your questions

The most common questions on this kind of case. Short answer first, detail after.

Yes, if the development goes beyond the state of the art and carries genuine technical risk, such as reliable real-time analysis under tight resource conditions. Merely applying finished models is not.

Development work on a new or substantially improved hardware-software system with an open outcome is eligible. In-house work usually has the largest lever.

The research allowance is calculated per fiscal year. For multi-year projects the funding adds up over the years, provided in-house work is captured year by year.

No. Even without external engagements the eligible base can be considerable if the company's own development work is cleanly captured as eligible personnel work.

Keep reading

More case studies

Biotechnology

Mechanics, control and biology funded as one interdisciplinary R&D project.