Prediction Of Street Lights In Metro Cities Using Time Series Analysis In Machine Learning Algorithm

Document Type : Primary Research paper


1 Associate Professor Department Of Information Technology Bannari Amman Institute Of Technology Sathyamangalam, India

2 Post Graduate in Software Engineering Department of Information Technology Bannari Amman Institute of Technology Sathyamangalam, India.


In the current scenario, the substitution of man power becomes liable in the
urban area where more sustainable projects have been developed to implement the smart
city environment. In this venture, the smart lighting in the streets are getting popular to
maintain the operations of the street lights. There are different other constraints in which
the smart lighting can be addressed in monitoring the usage of the resources of the
particular LED bulbs and replacement time period of the LED Bulbs placed in the
different location. A device designed to monitor the LED bulbs intensity and usage time of
the LED bulbs in the particular location will be sent through the Message Queuing
Telemetry Transport –SN (MQTT-SN) protocol to the IOT web environment. The
information is maintained in the real time database where time based autoregressive
integrated algorithm with moving average is used to predict the usage of the particular
bulb and alerts the replacement of the particular bulb before the bulb gets worn out. The
analysis is performed with respect to the life time of the bulb and
their usage in the particular environmental factors such as temperature and humidity at
the location.


Volume 12, Issue 3 - Serial Number 3
ICMMNT-2021 International Virtual Conference on Materials, Manufacturing and Nanotechnology, 30th June, 2021.
June 2021
Pages 1336-1345