program

A personalized caching scheme for mobile Location-based Social Networks
Authors: D. Tomaras, I. Boutsis, V. Kalogeraki (Presenter), D. Gunopulos

Drone Swarms for Urban Sensing

Abstract: Urban sensing refers to the collection of data from urban (city) environments, aggregating and learning from the data, then using the processed data in making near term decisions. Sensing traffic data by loop detectors and video cameras and using the data to modify traffic light delays, detecting and responding to noise pollution problems, and monitoring chemical pollution levels are all examples of urban sensing. Traditional urban sensing approaches suffer many drawbacks, including significant installation and maintenance costs, and poor coverage in rural areas. Drones an drone swarms can serve as a cheap and effective alternative to traditional approaches. We introduce some of the challenges of using drones for urban sensing, and the tools which can be used to solve them. We motivate the need for drone swarms in urban sensing as opposed to individual drones, and introduce the challenges and solutions for coordinating swarm actions. As a case in point we discuss the Q-learning implementation we used to attack the broadcast storm problem in drone swarm communication. In this talk, we will illustrate our ongoing work on adaptive data processing and dissemination for drone swarms in urban sensing.

Author and Presenter: Dmitri Arkhipov