What

The purpose of this project is to assist with the development and evaluation of a wireless sensor network-based system for monitoring large-scale urban infrastructure such as water supply and sewer networks. The work is in collaboration with a research group at the Civil Engineering Department at MIT led by Professor Andrew Whittle (sensor.mit.edu). The main objective is to bridge advances in hydraulic modeling, signal processing and sensor networks in order to demonstrate a prototype low-cost system for continuous monitoring, fault identification (leak detection) and optimal control (pressure optimization) in water distribution and sewer networks.

How

The monitoring is done by clusters of radio-equipped battery-powered sensors, and the collected data is relayed via a 2-hop network to an Internet-connected server (the cluster head). The data collection network is arranged into two tiers. The first tier consists of sensor nodes that have so far been interfaced with pressure transducers, hydrophones, pH probes and ultrasonic level sensors. The sensors forward the acquired data using bluetooth wireless radio to the cluster head. The cluster head controls the sampling regime of the sensor nodes. The two permitted modes are burst event monitoring (up to 1000 S/s) and long-term periodic monitoring. Burst event monitoring permits focusing the system on observation of a single short event, while the long-term sampling regime allows monitoring of the infrastructure in the long run. Finally, the cluster head transfers the data to a data collection center (via GPRS or WiFi).

The data center uses the acquired data for the state estimation and fault identification of the monitored system. For example, the data center can use water pressure variations to detect reflections of a pressure wave off of small cracks in the pipe to localize the position of imminent failures [1,2].

Goals

The work carried out within this project will concentrate on developing the software for transferring the data securely and reliably from the cluster head to the data collection center via GPRS in a dual-way communication. The work will also involve the development of software tools to discipline the oscillator of the cluster head in order to provide accurate time beacons for synchronizing the acquired data. Two techniques will be explored:

• using NTP over GPRS; and,
• using the 1 PPS (pulse-per-second) provided by GPS (Global Positioning System).

Progress Thus Far

We are in the process of setting up a prototype of the system that can diagnose extreme conditions that threaten urban infrastructure. The current progress of this project can be summarized as following:

• An initial field trial is underway in South Boston. The monitoring system was deployed on three measurement locations (3 cluster heads and 5 sensors) in collaboration with Boston Water and Sewer Commission on the 21st of December 2004. The data collection however was limited to a drive-by method. The second stage of this deployment is scheduled for the end of April which will aim to complete the remote data collection process controlled by the cluster heads via GPRS.
• A version of the software that runs on the cluster head that supports intermittent GPRS connectivity has been developed and is currently being tested.
• The accuracy of running NTP on the cluster head is being evaluated under various traffic conditions and levels of connection intermittency.
• A second generation of the Bluetooth communication software between cluster heads and cluster nodes is being developed by Intel Research.

The sensor nodes are deployed in hard-to-reach locations and have limited power and network connectivity. Hence, we anticipate a number of challenges related to dealing with networking, power management, time synchronization, and remote administration as the project develops.

References:

[1] Stoianov, I., Dellow, D., Maksimovic, C. and Graham, N.J.D. Field Validation of the Application of Hydraulic Transients for Leak Detection in Transmission Pipelines. In The Proceedings of CCWI 2003 Advances in Water Supply Management Conference, London, UK, September 2003.

[2] Stoianov, I., Maksimovic, C. and Graham, N.J.D. Designing a Continuous Monitoring System for Transmission Pipelines, In The Proceedings CCWI 2003 Advances in Water Supply Management Conference, London, UK, September 2003.