Keynotes
Keynote 1: The sound of health and fitness sensing
Cecilia Mascolo
University of Cambridge, UK
Abstract: Wearable devices are becoming pervasive in our lives, from smart watches measuring our physiology to wearables for the ear accompanying us in every run or virtual meeting. The monitoring of our health and fitness through sensors and wearables is also the focus of much research in the pervasive computing community. However, despite commodity devices being able to monitor our wellness and activity, many challenges still exist before truly scalable, trustworthy and affordable health and fitness monitoring becomes a reality. In this talk I will discuss where commercial systems have gotten to today and highlight the open challenges that these technologies still face before they can be trusted health measurement proxies. Namely, the ability to work in the wild and to cope with the variability of uses; the trade offs that we need to consider with respect to the sensitivity of the data and the use of constrained on device resources; the uncertainty of the prediction over the data. I will mostly use examples from my team's ongoing research on audio based machine learning for health, on-device machine learning and "hearable" sensing to explore the current and future path of pervasive computing for health and fitness.
Keynote 2: Climate Smart Computing: A Perspective
Shashi Shekhar
University of Minnesota, Twincities, MN, USA
Abstract: Climate change is a societal grand challenge and many nations have signed the Paris Agreement (2015) aiming for net-zero emission (a.k.a. Carbon neutrality) around 2050. The computing community can make many contributions. We can help improve climate resilience, lower emissions in other economic sectors (e.g., energy, transportation, agriculture) and accelerate absorption of greenhouse gasses (GHG) in nature. Also, we can reduce computing emissions (about 4% of total in 2020 from IoT, pervasive devices, networks, data centers, etc.). Further, we can provide new tools for advancing scientific understanding and for GHG monitoring, reporting and verification.
However, traditional computing methods face major challenges. First, computational models are approximations of the natural world, and it is important to reduce approximation error due to the high cost of errors besides speeding up computations. In addition, there are significant interactions and optimizing for a goal (e.g., mitigation) or a subsystem (e.g., food) may have unintended consequences for other goals (e.g., adaptation) or subsystems (e.g., water). Moreover, machine learning is overwhelmed due to non-stationarity (e.g., climate change), data paucity (e.g., rare climate events), high cost of ground truth collection, and the need to observe natural laws (e.g., conservation of mass). Furthermore, addressing climate often requires deep and sustained interdisciplinary collaboration.
This keynote shares a perspective on the climate-smart computing challenges and opportunities based on multi-decade scholarly activities such as the recent AI-CLIMATE, the National AI Research Institute for Climate-Land Interactions, Mitigation, Adaptation, Tradeoffs and Economy.