Residential Move-in Detection & Machine Learning - Junior Developers Singapore

Published on: Wednesday, 13 February 2019

Speaker: Chang Kai Wen, Intern, SP Digital

Full title: Residential Move-in Detection using Utilities Consumption Data & Machine Learning

This project is part of a bigger project which uses AI to improve the efficiency of metering to billing operations for SP Services. Meter readings are sent to the server in SP at every two months. When an electricity meter reading is sent back to the system, the reading will be checked against a system of rules. If the meter reading is too high, it might be caused by a faulty meter. Hence, technicians need to be sent to the premises to do on-site investigation. However high meter readings are also often due to someone moving into the premise, and therefore is a correct reading. This means the trip, which incurs manpower and other costs, is wasted.

Unfortunately, the system of rules implemented few years ago are unable to tell the difference between a legitimate high consumption reading or an actual faulty meter. My project is to build a predictive machine learning model to detect new move-in events. By integrating this model into the daily operational flow of a meter irregularity investigation, the operations team will be able to identify if the high consumption is due to a new move-in or not and reduce false positives.

Event Page: https://www.meetup.com/Junior-Developers-Singapore/events/258349143/

Produced by Engineers.SG

Recorded by: Wing Puah and Yeo Kheng Meng

Help us caption & translate this video!

https://amara.org/v/nb96/