In recent years, Structural Health Monitoring (SHM) has emerged as an important research area in civil engineering. To investigate both local and global damage criteria, a dense array of sensors is anticipated to be required for large civil engineering structures. Traditional centralized data acquisition systems do not provide a scalable approach, since the shear number of accompanying wires, fiber optic cables, or other physical transmission medium may be prohibitive. Wireless communications have the potential to significantly impact monitoring systems. To assist in dealing with the large amount of data that will be generated, on-board processing at the sensor allows a portion of the computation to be done locally on the sensor's embedded microprocessor. The Mica Mote platform, along with Tiny Operating System (TinyOS) developed at the University of California at Berkeley offers for the first time, an open hardware/software environment for broad smart sensing research. However, because the accelerometers on the existing sensor boards have both poor low frequency sensitivity and high noise density, their suitability for civil engineering applications is unclear. This paper presents results for a new sensor board employing an SD-1221 accelerometer, which overcomes many of the deficiencies of the existing sensor. Finally, a number of the challenges still remaining are identified.