Comparison of Statistical Analysis Methodologies for LED Indication Replacement Schedules

The application of Light Emitting Diode (LED) lighting systems has experienced significant growth in the transportation sector over the past 10 years. LED indication lifespans have significantly greater durations than previous technologies, however, uncertainties in the duration of LED devices have unearthed challenges in developing a replacement schedule of LED traffic indications. This research evaluates two methods to approximate the useful life of LED traffic indicators. Previous research in LED indications used a lateral regression analysis to determine the lifetime of LED indications. A new methodology is used to incorporate longitudinal regression analysis as well as account for additional factors outside previous research methods. The findings of this paper compare the results using the new methodology against previous research recommendations. A discussion on the impact of the updated results on transportation agencies' policies is also included.