ECSEL JU Project AI-TWILIGHT

In all lighting sectors, warranty and customisation are becoming key product differentiators. In addition, the integration of more electronics and sensors in lighting systems will change what we call ‘lighting’ today.

While the concepts of digitalisation and Industry 4.0 are progressing fast into the manufacturing world, front-end product design in the lighting industry still uses traditional simulation techniques, while the back-end struggles to use all the data generated by sensors. An innovative approach is to couple digital twins with Artificial Intelligence to offer unlimited possibilities to the “first build and then tweak” approach.

AI-TWILIGHT will merge the virtual and physical worlds to pave the way for innovations in fields where the European lighting industry is likely to be competitive. Self-learning digital twins of lighting systems (LED source, the driver of a lighting application) using AI and analytics techniques will be created and used as input for predicting performance and lifetime of autonomous product and infrastructure design and management. Tests will be carried out in selected application domains e.g., automotive, horticulture, general and street-lighting.

Translated to business goals, AI-TWILIGHT will result in 20% increase in the introduction of more customised and connected products while reducing the time to market by 30% and reducing by 25% the total cost of ownership of an AI-TWILIGHT powered system. The project will also promote extensive energy savings (of around 30-40%, depending on domain) by enabling optimally designed luminaires with higher efficacy and longer foreseen useful lifetime, as well as through intelligent control of lighting.