Marconi Society’s Celestini Programme recognises Indian students tackling air pollution and road safety; winning team develops Android app that estimates air pollution level from photographs.
A team from Bharti Vidyapeeth College of Engineering, Delhi has developed an Android app that uses smartphone camera images to estimate the Air Quality Index (AQI) levels in the user’s neighbourhood. The solution took the top prize recently in a contest organised in India by the Celestini Programme, supported by the Marconi Society. The Celestini Programme, named for the hill in Italy where Guglielmo Marconi conducted his first wireless transmission experiments, is run by winners of the Society’s annual Young Scholar Awards, who work with technical undergraduate students in developing countries, to use technology to create social and economic transformation in their communities.
The winning team of Tanmay Srivastava, Kanishk Jeet and Prerna Khanna, developed an inexpensive, portable and real-time air quality analytics app: Air Cognizer. The user uploads an image taken outdoors with half of the image covering the sky region. Using image processing techniques, features are extracted and the machine learning model estimates the AQI for the user’s location. The machine learning model is deployed on smartphones using Tensorflow Lite and ML Kit from Google. An Android app of the same name is available at Google Play.
Air Cognizer is simple to use and free —and will prove to be very useful for citizens in cities like Delhi, where air pollution is particularly acute now. The winning team receives a cash prize of $1,500.
In India, the Celestini Programme was started in 2017 in partnership with IIT Delhi by Aakanksha Chowdhery, machine learning engineer with Google AI, who was selected as a Marconi Young Scholar in 2012 for her work in high-speed last-mile internet connectivity. IIT Delhi partners include Brejesh Lall and Prerana Mukherjee.
The Celestini Programme has hosted 14 students in India so far. In 2018, the second year of the Program in India, three teams from among 100-plus applicants were selected to work during the summer at IIT Delhi on problems related to air pollution and road safety in New Delhi.
The second prize went to the team of Divyam Madaan and Radhika Dua, from UIET Chandigarh, Punjab University. They prototyped a website that forecasts air pollution levels in Delhi over the next 24 hours using advanced machine learning techniques such as LSTMs to predict the major pollutant and its cause (for example, road traffic, industry emissions, or agricultural wastes) in every location based on historical data. The website prototyped by the students, updates in real-time using Google Cloud platform and Cloud ML engine.
The third team, also from Bharti Vidyapeeth College of Engineering, included Sidharth Talia, Nikunj Agarwal and Samarjeet Kaur. They prototyped a low-latency platform to transmit vehicle-to-vehicle alerts about potential road safety hazards or collisions using computer vision techniques on Raspberry Pi and XBee radio modules.