GlassEye: Winning Innovate UK Bilateral Funding for AI-Powered Glass Inspection

At a glance

The challenge

Advanced high-tech glass sits inside semiconductor lithography photomasks, photonics, displays and PV panels, and these applications demand high-throughput, zero-defect production. Yet defects such as bubbles, inclusions and blisters often go unnoticed until the glass has been polished, cut and bonded, at which point it is non-recyclable. Production lines shut down and entire batches are discarded, costing manufacturers millions.

The funding route added its own difficulty: a bilateral competition means two funding bodies, two rule books and one shared work plan. The UK side had to satisfy Innovate UK's assessors on need, innovation, team, market and value for money, while the German partner's costs ran through ZIM in euros with their own conversion and reporting rules. Getting one coherent, winning story across both was the job.

The solution

GlassEye is a light-field (plenoptic) camera system integrated directly into the production line. It captures high-resolution 2D texture and 3D shape in a single shot and can see beneath surface layers where light scatters. An AI inspection platform built on hybrid multimodal deep learning models detects, localises and classifies defects in real time, then feeds the results back to the line: discard, re-melt, or adjust upstream process parameters.

What we did

1. Sharpening the concept

We worked with the partners to frame GlassEye around quantified need: the $2.53 billion surface vision and inspection market, a high-tech glass market heading towards $70 billion by 2030, and the concrete costs of manual inspection. Every claim in the need and innovation sections is tied to a number and a source, from Forbes' estimate of up to $50 billion in annual unplanned downtime costs to the UK's daily landfill figures.

2. Structuring a two-country bid

We aligned the UK and German halves into one work plan with clear roles: IVY TECH leading integration and project management, Brunel University London building the machine learning pipeline, Glass Technology Services running glass trials and industry dissemination through the British Glass Manufacturers Confederation, and Raytrix developing the camera sensor. Budgets were structured to each funder's rules: £632,891 UK costs with a 78% Innovate UK grant rate, and €375,740 German costs through ZIM.

3. Making the business case

Assessors fund returns, not just technology. The proposal set out a five-year commercialisation plan: a £85,000 unit price, market entry through GTS's UK industry access and Raytrix's sales network in Germany, Korea and the USA, and projected revenues of £39.5 million with £31.1 million profit, a 22.9:1 return on grant investment and 82 new jobs in the five years after the project.

4. Impact beyond the balance sheet

The bid quantified environmental gains (glass melting consumes around 1.6 kWh and 0.33 kgCO2e per kilogram; cutting a 15% reject rate across the UK's 850,000 tonnes of annual flat glass production saves both), and aligned the project with the UK National Semiconductor Strategy and Industry 4.0 priorities.

The result

Over 24 months the consortium will take the integrated camera and AI platform through lab trials with real production glass to an in-line demonstration, on the path to a commercial product for glass manufacturers across the UK, Germany and beyond.

Why this matters for your organisation

Bilateral and multi-funder competitions are among the trickiest applications to get right: two sets of rules, two evaluation cultures, one story. GlassEye shows the approach that works: quantify the need, give every partner a clear role, and make the commercial and environmental returns impossible to miss.

If you are preparing an Innovate UK, Eureka, bilateral or Horizon Europe application, we can support concept development, consortium structuring, proposal writing, evaluator-style review and full submission. And if a project is not competitive, we will tell you before you spend money on it.