Building on previous work at Georgia Tech by Dr. Russ Clark and the Smart Sea Level Sensor (SSLS) project, the WAVES project aims to extend the range of feasible locations for low-cost waterlevel monitoring sensors by using low-cost satellite modems to automatically send back data. While the SSLS project has achieved much early success in both providing real-time safety information to Savannah residents and measuring longterm changes in sea level, the sensors rely on existing Internet infrastructure to upload their data. This limitation makes deployments in remote locations like Sapelo Island impossible even though they are some of the most important places to measure. Through funding provided by the Georgia Tech Center for the Development and Applications of the Internet of Things Technologies (CDAIT), I am leading a team of Georgia Tech undergrad and graduate students in developing a water-level monitoring system that can be deployed and automatically upload data from anywhere regardless of existing Internet infrastructure.
In designing the new sensors, we acquired some Swarm modems for around $100 each and are replacing the existing LoRa communications backhaul with these modems. For a $5 a month data plan, each device can send 750 data packets of 192 bytes each, which presents significant challenges in bundling and formatting data to due to the small amount of information we can send. We are applying additional engineering effort into:
- Optimizing when to transmit data since the satellite constellation is sparse and connectivity is transient.
- Figuring out relevant parameters and optimal locations for robust satellite connectivity, especially with concern paid towards the tree cover and other environmental factors that can limit RF communications.
- Reducing the energy consumed by our battery-operated system and exploring energy harvesting.
- Ruggedizing the sensors for extended field deployments. We plan to conduct field tests with our final prototype in Atlanta and Savannah before exploring expansions into other areas in which it was previously infeasible to automatically collect data.