Reliability of New Light Quality Metrics: Challenges and Opportunities
Together with high energy saving potentials, LED reliability is considered as one of the key advantages of this technology for artificial lighting . The promise of lighting systems with high reliability leads to various advantages such as lower maintenance needs, long term cost savings and a reduction of the impact of production on the environment.
With the growing maturity of LEDs and the increase control over their spectrum, new applications focusing on the effect of light beyond its conventional illumination function are emerging . As a result, the quantity of LED based lighting solutions for horticulture, agriculture or wildlife care is constantly increasing . In a wider scale, the impact of light on humans, including its impact on the circadian rhythms, is more and more considered and is foreseen as one of the major trend for indoor lighting in the coming years . Those “beyond lighting” applications require the implementation of new metrics (e.g Cs/CLA or ML for circadian lighting). Reliability is defined as the ability of a component or system to fulfill its required functions in a certain environment for a specific period. Therefore, for these applications, the lighting system function is not only illumination but can also be composed of other requirements (e.g. a certain plant growth speed, a certain color saturation level). For those applications, reliability of the lighting systems is no longer defined only with a flux maintenance. As a result, the maintenance of those new light quality metrics must be analyzed and modeled. However, as those metrics are directly linked to the spectrum of the light source, reliability models based on its variations have to be developed. Due to different LED technologies, especially regarding light conversion, and to the amount of possible stress factors, spectrum reliability models are challenging to build.
In this paper, the impact of spectrum ageing on different light quality parameters will be presented (including TM30 and circadian parameters). Based on resent results (composed of simulations and ageing data) the influence of technologies on the maintenance of those parameters will be exposed highlighting the advantages and limits of the different technological approaches (e.g. multiple phosphors vs. n channels). In a third part, the challenges of modelling the light quality parameters maintenance over time will be discussed. Finally, recent results obtained using innovative modeling tools based on artificial intelligence and machine learning will be presented underlining a solution to the modeling of light quality parameters maintenance.