In the past 4 years we have witnessed a change in quality and reliability to make the marked introduction of solid-state lighting (SSL) successful. LED penetration levels have reached values of 10–30%, depending on the application. The number and variety of LED packages and, thus
...
In the past 4 years we have witnessed a change in quality and reliability to make the marked introduction of solid-state lighting (SSL) successful. LED penetration levels have reached values of 10–30%, depending on the application. The number and variety of LED packages and, thus, associated LED-based products, have significantly increased in the past years. Consequently, new processes and new materials are introduced which will introduce a new series of new and unknown failure modes in SSL products. The understanding of these failure modes is better understood, and the number has grown to beyond 50. The fingerprint is changing to failures that are due to interactions between components. First exercises with system level acceleration tests are presented, but it is important to derive acceleration models for these tests. Advanced reliability prediction capabilities are needed including algorithms and tools that couple the multi-physic and multiscale behavior of the SSL failure modes. The shift toward services will force the lighting industry to develop these capabilities in order to better address lifetime and reliability. Connected lighting will bring big data from live connections that can be used to determine the degradation level of the system. Both trends, service and connected, will bring yet another huge change in mind-set in the lighting industry when in concerns reliability: detailed understanding of failure mechanisms, usage scenarios, technology, and design will come together.@en