Home » Project 3
Project 3
Project Outcomes
In order to achieve the objectives of a Geophone that could detect -2 on the Richter scale we applied adaptive filtering algorithms that adjust based on real-time noise levels, such as band-pass filters to isolate seismic frequencies while eliminating environmental and electronic noise. These algorithms actively reduce interference from non-seismic sources like wind, traffic, or machinery, similar to noise-canceling in audio devices but adapted for seismic frequencies. The team experimented with electromagnetic shielding solutions to ensure the purest signal without interference. In post-processing, wavelet transforms and machine learning were used to further refine the seismic signal, isolating the minute events from background clutter with high precision.
A number of solutions were attempted in developing a suitable prototype. The design included selecting high-gain antennas and optimizing signal strength to penetrate geophone burial housings while maintaining a strong connection. Special attention was given to material properties of the housing to minimize signal attenuation, ensuring the 2.4GHz WiFi signal could travel up to 30 meters, even when the geophones were not visible from the surface. A low-power management system was designed where geophones enter a sleep mode to conserve battery. The drone, upon approach, wirelessly activates the geophone’s WiFi module to enable the data transfer. This ensures that the WiFi is only active during data download periods, conserving energy. A GPS-based locationing system for each geophone was integrated into a database. The drone autonomously follows pre-programmed flight plans to reach the geophones based on their coordinates, downloading data logs on a periodic schedule. Flight plans are generated based on the most efficient routes, considering terrain and battery life. Each geophone is equipped with a GPS tracking device to monitor its position. A backup monitoring system sends alerts to the base station if a geophone is disturbed or moved unexpectedly. If this occurs, a drone-based rapid response system is triggered, where a drone flies to the affected geophone to assess the situation and either recover the data or attempt immediate repairs if possible.
A robust SQL-based database was designed to handle ultra-large datasets generated by passive seismic tomography. Cloud-hosting services were used to store and process the massive data, ensuring scalability and real-time access. The first prototype geophone was tested in the field to capture seismic data, comparing performance against conventional seismic equipment. Data integrity and sensitivity were key focus areas. The drone periodically collected data from the geophones in the field, ensuring efficient data transmission without manual intervention. The data collected by the drone system was integrated into the virtual seismograph system. The data was processed, and initial seismic results were compared to conventional equipment, showing comparable performance. Iterative changes based on real-world conditions were applied, such as improving drone battery life and refining geophone sensitivity. The prototype demonstrated the potential for mass deployment with future improvements focused on scalability and robustness in challenging terrains.