Skip to main content.
Advanced search >
<< Back to previous page Print

<< Tuesday, May 14, 2013 >>

Remind me

Tell a friend

Add to my Google calendar (bCal)

Download to my calendar

Bookmark and ShareShare

[Dissertation Talk] Wireless weigh-in-motion: using road vibrations to estimate truck weight

Seminar: Departmental | May 14 | 2-3 p.m. | Cory Hall, Hughes Room/400

Ravneet (Ronnie) Bajwa, UC Berkeley

Electrical Engineering and Computer Sciences (EECS)

Truck weight data is used to make important decisions concerning pavement design, road maintenance, and transportation policy at both the state and nation levels. Widespread use of weigh stations and weigh-in-motion (WIM) stations is recommended by experts but high installation and maintenance costs limit current deployments. In this talk, I will present a new cost-effective alternative: a wireless sensor network that estimates the weight of moving trucks by measuring the corresponding road vibrations. The system was tested using real trucks on a highway under varying environmental conditions. The wireless WIM passed the standard accuracy requirements (± 10% error) and outperformed a nearby government-operated WIM station. The following topics will be covered in the talk:

• Design of a low-noise wireless vibration sensor. The sensor has a resolution of 400 µg in noisy highway environment and is immune to any interference due to sound.

• A new pavement-vehicle interaction model that relates a vehicle’s load to the pavement acceleration (and displacement), temperature, and vehicle speed.

• Automatic vehicle classification. Vibration data is used to detect axles and vehicles are classified based on their axle count and inter-axle spacings.

• Load estimation. Vibration, temperature, and vehicle speed data are processed to estimate individual axle loads.

• Displacement estimation. Pavement deflection, as opposed to acceleration, is used by engineers for evaluating pavement performance. We present and compare algorithms to estimate pavement displacement from noisy acceleration measurements.

• Distributed computation and sensor-level data compression techniques to minimize sensor computation and data transmission for a longer lifetime.