Making Roads Safer by Making Drivers Better

Hari Balakrishnan

The world's roads see over 50 million injuries and 1.25 million fatalities every year; road accidents are now the leading cause of death among people between the ages of 15 and 30. This talk will describe how mobile sensing, large-scale data analytics, and applied machine learning can improve road safety by making people better drivers. I'll discuss several research and engineering challenges in achieving this goal, as well as learnings from successful commercial deployments in multiple countries. Interesting problems include inferring vehicular dynamics and trajectories from noisy sensor data; activity classification; detecting and discouraging distracted driving; designing good incentives, feedback, and "safety challenge" games; back-end scaling of a workload consisting of resource-intensive processing and interactive social features; and the design of new sensing platforms.


Hari Balakrishnan is the Fujitsu Professor of Computer Science and a Director of MIT's Center for Wireless Networks and Mobile Computing. His research is in networked computer systems, with current interests in networking, data management, and sensing for a world of "truly mobile" devices connected to cloud services running in large datacenters. He and Sam Madden led the CarTel project at MIT, which made several pioneering contributions in the area of vehicular sensor computing and networking. Building on this research, they started Cambridge Mobile Telematics (CMT), a company that develops mobile sensing, inferencing, and data analytics technologies to change driver behavior and make roads safer around the world.

How I Learned to Stop Worrying and Love Transactions

George R. Hernandez


George Hernandez joined PNNL in 2009 and works in the Advanced Building Controls group. Mr. Hernandez is a Staff Scientist and senior demand side management professional with innovative and detail-oriented knowledge to develop and produce successful programs that deliver products and services to the commercial and industrial energy marketplace. Mr. Hernandez is distinguished by exceptional execution skills that enable efficient concept to product delivery. Currently, Mr. Hernandez is on detail at the Department of Energy as a Technical Advisor specifically in the area of Sensor and Controls. While at DOE, he has co authored the High Performance RTU Challenge, the Buildings Performance Database, the Low Cost Wireless Metering Challenge, Energy Information Handbook, the Portable Sensor Suitcase, Open Source Small Building Control System, and the Transactional Network project. Mr. Hernandez has extensive knowledge, skills, and capabilities derived from a substantial career in demand side utility management across a wide variety of commercial and industrial sectors and utilities as both a corporate employee and an independent consultant. Mr. Hernandez received his BS in Mechanical Engineering from California State University and his Masters in Mechanical Engineering from The University of California at Berkeley. He is a Licensed Professional Engineer (PE) by the State of California.