Keynotes

Moonshot for Mental Health

ABSTRACT: Mental illness touches nearly every family. It negatively affects mood, behavior and functioning. It can come with tremendous personal cost. Consider the first year college student who experiences their first depressive episode on campus after feeling anxious, isolated and suicidal. Or the outpatient living with schizophrenia who endures severe hardship, such as, homelessness, victimization and incarceration. Over the last decade, we have seen the rise of digital mental health sensing and intervention technology based on phones, wearables and AI/machine learning. In my talk, I will discuss mobile sensing technology for the assessment, prediction and intervention of depression and anxiety for college students. I will argue that mobile and AI technology is poised to radically change how we diagnose, manage and treat student mental health on college campuses – in a nutshell, a moonshot for mental health.

BIO: Andrew T. Campbell is the Albert Bradley 1915 Third Century Professor in Computer Science at Dartmouth College. He is known for his pioneering work in mobile phone sensing. Some of the activity inference algorithms developed by his group are now common in all smartphones. Before joining Dartmouth, he was a tenured professor of electrical engineering at Columbia University working on mobile computing and communications. He has been a visiting professor at CMU Rwanda, University of Salamanca, Cambridge University and University College London. At Google he worked on cardiovascular health as a member of the Android group and later as a visiting research scientist at Verily Life Sciences working on mental health sensing. His work has received a number of awards (e.g., ACM UbiComp 2022 10-year Impact Award, the ACM SIGMOBILE 2019 Test of Time Paper Award where his group “pioneered applying machine learning across local devices and servers”) and has been covered widely by the popular press (New York Times, Financial Times, Economist), TV (BBC, CBS) and radio (NPR).

---

A Low-Cost LoRaWAN sensor network for analyzing urban heat islands

ABSTRACT: The climate change is leading to higher temperatures in many regions of the world, even in central Europe. Another observation is that there is ongoing soil sealing and construction in cities creating urban heat islands with significant increase in temperatures during days and nights. However, there are countermeasures such as the deployment of plants and the removal of concrete. Geography scientists at our university are analyzing the temperatures on a fine grained level in the city of Bern. In a collaboration we designed, implemented, and deployed a sensor and interconnected the sensors via the LoRaWAN-based Helium network. The sensor must be cheap due to the large number of devices to be deployed but also accurate in order to measure small temperature differences within a city. Moreover, it must be autonomous and cannot be connected to electricity. A special feature of the sensor is a ventilator pushing accumulated air out of the sensor to make measurements more accurate. However, ventilation must be supported by a small solar cell. We discuss the design details of the sensor and provide first measurement results. We also discuss the Helium network that is used for providing the connectivity. We further discuss possible improvements of the system using machine learning techniques.

BIO: Dr. Torsten Braun is currently director at the Institute of Computer Science, University of Bern, where he has been a full professor since 1998. He got the Ph.D. degree from University of Karlsruhe (Germany) in 1993. From 1994 to 1995, he was a guest scientist at INRIA Sophia-Antipolis (France). From 1995 to 1997, he worked at the IBM European Networking Centre Heidelberg (Germany) as a project leader and senior consultant. He has been a vice president of the SWITCH (Swiss Research and Education Network Provider) Foundation from 2011 to 2019. He has been a Director of the Institute of Computer Science and Applied Mathematics at University of Bern between 2007 and 2011, and from 2019 to 2021

---

Efficient Collaborative Optimization and Execution of Neural Models for Real-Time Applications

ABSTRACT: Neural analysis is arising as a central component of a broad range of applications. However, the complexity of the tasks and models, the diversity of operating contexts and lack of data, as well as channel and computing resource scarcity challenge the effective deployment of neural logics in many scenarios of great societal interest. In this talk, I will discuss two use-cases of great societal relevance: mobile healthcare and vehicular autonomy. I will present a set of solutions enabling online optimization, effective distributed execution and runtime adaptation of neural models in challenging applications scenarios.

BIO: Marco Levoratois an Associate Professor in the Computer Science department at UC Irvine. He completed the PhD in Electrical Engineering at the University of Padova, Italy, in 2009. Between 2010 and 2012, he was a postdoctoral researcher with a joint affiliation at Stanford and the University of Southern California. His research interests are focused on distributed computing over unreliable wireless systems, especially for autonomous vehicles and healthcare systems. His work received the best paper award at IEEE GLOBECOM (2012). In 2016 and 2019, he received the UC Hellman Foundation Award and the Dean mid-career research award, respectively. His research is funded by the National Science Foundation, the Department of Defense, Intel and Cisco. In 2020-2021, he was the vice chair of the IEEE Technical Committee on Smart Grid Communications. He serves in the TPC of IEEE Infocom, IEEE Secon, and ACM MobiHoc, is an associate editor of the IEEE Transactions on Communications, and was part of the organizing committee of several IEEE and ACM conferences, including IEEE Secon 2022 and 2017, ACM MobiSys 2015 and ACM MobiCom 2015 and 2014. In 2022, he gave the keynote speech at IEEE HealthCom.