This work presents a reliability analysis approach for connected outdoor lighting systems using cloud-derived field data. Traditional in-house reliability models often struggled with validation due to limited and imprecise failure and usage data, particularly concerning operational burning hours. With Signify’s installed base of over 100 million connected light points, large-scale cloud data from luminaire controllers now enables more accurate reliability assessments. The collected data includes detailed operational states and failure information transmitted via gateways to the cloud, allowing improved estimation of failures in time and validation of reliability predictions. The study demonstrates how cloud analytics can enhance forecasting accuracy and discusses the challenges of cleaning and interpreting datasets from connected lighting systems. These challenges are rooted in product- and business-specific characteristics that differ significantly from other industries, such as automotive or industrial equipment.
Speaker: Dr. Piet Watté – Signify Research Eindhoven