Sign In

Communications of the ACM

ACM TechNews

Predicting Breakdowns


View as: Print Mobile App Share: Send by email Share on reddit Share on StumbleUpon Share on Hacker News Share on Tweeter Share on Facebook

Researchers at the Rochester Institute of Technology (RIT) and Lockheed Martin are working on a sensor-based monitoring system that assesses the health of a vehicle and alerts the driver to potential problems. The system uses a network of embedded smart sensors that are placed near problem-prone components. Information from the sensors is wirelessly transmitted to a central command center for analysis. Nabil Nasr, director of RIT's Center for Integrated Manufacturing, says the system goes beyond existing technology by predicting the future health or failures of vehicles. The project is part of a $150 million contract between Lockheed Martin and the U.S. Marine Corps, which will equip 12,000 military vehicles with the technology so the health of a vehicle can be quickly assessed before it is sent on a mission. The system includes both standard sensors for temperature and vibration monitoring, as well as customized smart sensors for monitoring the vehicle. Nasr says the researchers have developed sophisticated software for analyzing the data produced by the sensors.

"The algorithms are extremely valuable because they help us build a model of predictive and condition-based maintenance, so we can predict failures before they occur, and we can make determinations about service based on the actual conditions of the equipment," says Randy Weaver, at the Rochester Genesee Regional Transportation Authority, which has been testing the technology for use in public transit systems.

From Technology Review
View Full Article


 

No entries found