SprAI

AI-powered concept, aiming to revolutionize gelcoat application by using advanced video detection to prevent blockages at the spray tip in real-time

The Challenger main goal is to develop an AI model which is going to receive a video stream of the spraying process in real time, and it will be able to successfully detect the Gelcoat spray, as well as accurately judge the correctness of the spray pattern.

When an anomaly in the spraying pattern is detected, the designed system should immediately start a cleaning process and perform test shots to identify if the problem is solved. Simultaneously, the expert engineer should be notified about the problem.

An important aspect of this problem is that the time window to detect and resolve a bad spraying pattern is very limited.

Manufacturing

Spraying Process

Innovative solutions to address the problem

The Challenger considers an AI expert model, which will be able to conduct Image recognition, in the context of machine vision, of the spray pattern and that will then be trained to decide over the quality of the pattern, to be a state-of-the -art solution to our need.

In more details, two RGB cameras will be placed inside the painting booth to monitor the whole painting procedure in real time. The next step will be to train the AI model to successfully recognize the gelcoat spray pattern. This is going to be a challenge because while spraying a mist is created around the spraying area which may lead to noisy data.

After successfully identifying the spray pattern the AI model will have to decide on the quality of the pattern.

TRYGONS S.A.

TRYGONS is a privately owned and innovative composite parts manufacturer, mainly active in the marine and automotive sectors. Leveraging its internal Research & Development capabilities, the company employs cutting-edge technologies to design and fabricate intricate reinforced parts of medium to large sizes, primarily catering to high production volumes.

The organization offers a comprehensive range of services, including composite part manufacturing, collaborative product development, encompassing initial conceptualization, detailed design, and rigorous testing, as well as material and product certification. TRYGONS ensures end-to-end support, from production to continuous improvement throughout the product’s lifecycle.

Recently, the company has initiated the use of AI for the improvement of its manufacturing processes. We are particularly interested in the digitization of our processes, for reasons of sustainability, flexibility, and overall efficiency.

Website

SprAI

The challenge we aim to address revolves around automating the gelcoat application process. The primary obstacle to automating this process lies in the frequent occurrence of spray tip blockages, which significantly impede the consistency of the layup thickness and overall product quality. A blockage of the spray tip leads to the production of heavily defective items, necessitating extensive repairs or direct scrapping.

Spray blockage represents a prevalent issue encountered within the Glass Fiber Reinforced Plastics (GFRP) sector, particularly in gelcoat application, a widely adopted manufacturing process for producing large composite parts like midsize to large yachts, automobile cabins, and wind blades.

For manufacturing companies, these blockages have direct implications, including increased production of defective products, escalated repair costs, material wastage, amplified environmental impact, and reduced operational efficiency.

Presently, the responsibility of assessing the correctness of the spraying pattern lies with the operator initiating the painting process, relying on their experience to evaluate the test shots. Errors in the spraying pattern are indiscernible during the painting process, with only blockages exceeding 90% detectable through a pump speed sensor, as substantial blockages lead to alterations in the gelcoat machine’s pumping speed. Unfortunately, by that point, irreversible damage has typically occurred.

Our motivation to address this problem encompasses several facets. Primarily, a significant economic impact is expected for manufacturers within our sector upon implementing this solution. Concurrently, it offers an avenue to reduce the environmental footprint by minimizing material waste. Lastly, it represents a substantial leap toward automation, particularly for medium-sized industries operating within the sector, as the gelcoat application process traditionally relies heavily on manual labour.

This Challenge is not eligible for the InnoBuyer programme co-creation phase.