Sompo Japan Nipponkoa Insurance, Accenture and a Japanese cab company will work jointly to develop a deep learning algorithm designed to help better understand driving behavior and improve safety in Japan.
Their collective goal: to reduce the number of accidents, boost taxi driver training and increase the use of safe drivers, all of whom could end up using personalized safety instructions as part of their jobs.
The plan is for Sompo Japan Nipponkoa Insurance to collect data from connected devices installed in Daiichi Kotsu Sangyo taxis in Japan. This will include camera related images and telemetry tools that record trip data. Consenting taxi drivers will also use wearable devices to collect biometric information such as heart rates
Once in hand, as part of an ongoing strategic relationship between Accenture and Intel, the data will be processed securely and anonymously using Intel’s Internet of Things Platform Reference Architecture technology. Then, the data gets uploaded to the cloud where it will be securely stored and analyzed.
The end goal here is for Accenture to use the data to develop an algorithm that automatically assesses accident risk for each driver. This is done by collating and analyzing images, biometrics and vehicle data involving speed and driving behavior. Another concept behind this: deep learning, an emerging advanced analytics technique, which Accenture/Sompo Japan Nipponkoa Insurance said will be key to the data platform.
All parties involved participated in an initial proof of concept experiment in March 2017 involving data from 100 taxis and 100 drivers. The companies said that the deep-learning algorithm developed intelligence that spotted signs of drivers’ drowsiness and near-miss accidents, all based on heart-rate changes and riving behavior.
Longer term, the ability to analyze images on a large, commercial scale is still under development. Accenture said it will continue creating new intelligence toward this goal by using the latest analytics technologies.
Sources: Accenture/Sompo Japan Nipponkoa Insurance, Daiichi Kotsu Sangyo