Design and development of smart sensing systems for a smart city (water, air and pedestrian)
Design and development of low-cost, low-power sensors and sensing systems are active research areas in building the effortless smart city. This dissertation proposes novel compact systems applicable to a smart city scenario. Firstly, a pedestrian counting and environmental monitoring sensor node has been developed to count the number of pedestrians and their travel direction and provide ambient parameters (temperature, humidity, pressure, CO2, and TVOC). The novelty of this work is improving the accuracy by innovative system design and developing intelligent algorithms even though using off the shelf devices. The importance of developing selective sensors to monitor air quality has been realised while working with commercial sensors. Hence, a graphene oxide (GO)-coated planar interdigital sensor has been developed to induce selectivity towards CO2. The sensor detects a wide range of CO2 concentrations from 400 ~ 4000 ppm. The novelty of this work lies in compensating the temperature and humidity effect to make it applicable in actual environmental conditions and measure CO2 concentrations with more than 95% accuracy. Moreover, Reduced Graphene Oxide and Aluminum-doped zinc oxide (rGO/Al-ZnO) and Reduced Graphene Oxide and Tin oxide (rGO/SnO2) nanocomposites have been synthesised for selective detection of Acetone (C3H6O) and Ethylene (C2H2), respectively. The sensors detect and differentiate acetone and ethylene concentrations from 10 ppb to 300 ppb. The significant contributions of this research are improving the response and recovery time, detection limit, and developing an intelligent sensing system to monitor air quality using these sensors. Water is another essential aspect of a smart city and smart agriculture. Hence, finally, a flexible Multi-Walled Carbon Nanotubes (MWCNTs)/Polydimethylsiloxane (PDMS) based sensor has been fabricated for nitrate (0.1~25 ppm), phosphate (0.1~25 ppm), calcium (1~200 ppm), magnesium (1~200 ppm), and pH (1.3~12.4) detection in water. A novelty of this work is utilising a machine-learning algorithm to classify the data and predict the amount of water quality parameters using only one sensor and without utilising selective coating. Additionally, a smart autonomous and portable system has been designed for real-time water quality monitoring from any remote location. These systems and sensors provide an opportunity for decision-makers to test assumptions and strategies and provide a healthy and safe environment for all of us.