Advanced search
Start date
Betweenand

Advanced solution for drivers identification and evaluation by means of machine learning in large scale and connected solar tracking

Abstract

Cobli aims to research, develop and market electronic devices and computer techniques to connect and analyze in real time, motor vehicles and their drivers, to ubiquitously, enabling intelligent and safe use of this resource; that is, maximizing the benefits and minimizing their costs and risks, both in the preservation of human life and the environment. The results obtained by Cobli so far are tangible and encouraging. This research project aims to improve the drivers identification and evaluation algorithms through advanced machine learning techniques on data from solar trackers connected to the Internet (GPS and acceleration in three axes).The objectives will be achieved with optimizations in data processing techniques, particularly in predictive algorithms and processing in a distributed manner, to enable multiple drivers analysis, and increase its accuracy, as well as the collection of these same data, enabling commercial application for insurance companies and fleet managers.The solar tracking device is also an innovative element of the proposal as it will allow easy installation even in vehicles without standard interface for access to the vehicle's on-board computer data. The team has already developed a successful proof of concept identification and evaluation software with field testing and approval market companies.In 2015, he developed a device prototype solar tracker with GPS and accelerometer. The device concept was validated successfully bench, but the company was faced with technological challenges to withstand inhospitable conditions such as high temperatures, vibration, wind gusts, dust, particulates in the air and impact as well as management optimization energy and the list of components to reduce the cost of the solution to the scale manufacturing, which also comprise objectives of this research. (AU)

Articles published in Agência FAPESP Newsletter about the research grant:
Startup uses artificial intelligence to analyze vehicle driver behavior  
Inteligencia artificial para el análisis del comportamiento al volante 
Articles published in Pesquisa FAPESP Magazine about the research grant:
Fertile ground for artificial intelligence 
Articles published in Pesquisa para Inovação FAPESP about research grant:
Startup uses artificial intelligence to analyze vehicle driver behavior