TY - GEN
T1 - Autonomous River Boat Sensor Platform
T2 - 2023 IEEE World AI IoT Congress, AIIoT 2023
AU - Maguire, Tara
AU - Meehan, Kevin
N1 - Publisher Copyright:
© 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - Climate change is changing landscapes and technology is racing to keep up. Government agencies are struggling to fund research and monitoring projects throughout the world, and this has allowed the opportunity for citizen scientists to get involved in monitoring the impacts of climate change. This research outlines a relatively low-cost prototype, made up of readily available technology, like household or commercial sensors and open-source software. This prototype can be adapted, for different sensors and monitoring requirements, to carry out monitoring along river stretches in an intelligent way. Sensor fusion will be used to maximise the information gained from the sensors. PCA analysis shows that testing parameters support the inclusion of four dimensions at 87% variance, but six will capture 97%. Testing datasets are small and further tests will clarify this in the future. Key to this prototype will be an embedded Random Forest model, trained on large water quality datasets, with an F-score of 0.85 and capable of dictating navigation parameters depending on the data received from the on-board water quality sensors in real-time. The target of collecting data from an array of sensors and using the data to control an autonomous vehicle has been tentatively achieved and future directions could be for greater sensor fusion and developing the prototype for waypoint following.
AB - Climate change is changing landscapes and technology is racing to keep up. Government agencies are struggling to fund research and monitoring projects throughout the world, and this has allowed the opportunity for citizen scientists to get involved in monitoring the impacts of climate change. This research outlines a relatively low-cost prototype, made up of readily available technology, like household or commercial sensors and open-source software. This prototype can be adapted, for different sensors and monitoring requirements, to carry out monitoring along river stretches in an intelligent way. Sensor fusion will be used to maximise the information gained from the sensors. PCA analysis shows that testing parameters support the inclusion of four dimensions at 87% variance, but six will capture 97%. Testing datasets are small and further tests will clarify this in the future. Key to this prototype will be an embedded Random Forest model, trained on large water quality datasets, with an F-score of 0.85 and capable of dictating navigation parameters depending on the data received from the on-board water quality sensors in real-time. The target of collecting data from an array of sensors and using the data to control an autonomous vehicle has been tentatively achieved and future directions could be for greater sensor fusion and developing the prototype for waypoint following.
KW - AI
KW - Arduino
KW - Autonomous River Monitoring
KW - Random Forest
KW - Raspberry Pi
KW - Sensor Fusion
UR - http://www.scopus.com/inward/record.url?scp=85166623180&partnerID=8YFLogxK
U2 - 10.1109/AIIoT58121.2023.10174488
DO - 10.1109/AIIoT58121.2023.10174488
M3 - Conference contribution
AN - SCOPUS:85166623180
T3 - 2023 IEEE World AI IoT Congress, AIIoT 2023
SP - 554
EP - 560
BT - 2023 IEEE World AI IoT Congress, AIIoT 2023
A2 - Chakrabarti, Satyajit
A2 - Paul, Rajashree
PB - Institute of Electrical and Electronics Engineers Inc.
Y2 - 7 June 2023 through 10 June 2023
ER -