Visiting Cambridge University

Over the last couple of days, I’ve been visiting Cecilia Mascolo and Alastair Beresford at Cambridge University, Fahim Kawsar at Nokia Bell Labs, and Nic Lane at Samsung AI Lab in Cambridge.

Initially I was here to give a talk in the Centre for Mobile, Wearable Systems and Augmented Intelligence on “A Decade of Ubiquitous Computing Research in Mental Health”. The abstract and the slides are available below.

I had a chance to talk informally about what people are doing now and present what we’re doing in CACHET, especially on the CARP Mobile Sensing framework. From Nokia Bell Labs I got a bunch of their new in-ear sensing platform – eSense – and I’m looking forward to integrate these with CARP so that we can start rolling out new sensing capabilities in our Ubicomp research.

Figure 1: eSense open wearable platform with audio, motion, and BLE sensing powered by a CSR processor and a 40mAH battery (picture from the MobiSys 2018 demo paper).

See the IEEE Pervasive Computing paper on the eSense platform for details.


This year it is ten years since smartphones became widely available as an open platform and have since then been used for creating novel personalised health applications. From the very beginning, there has been an interest in exploiting the advantages of mobile and wearable technologies in mental health to unobtrusively sense and analyse human behaviour, assess and predict mental health status, and to deliver feedback and intervention when needed.

In this talk, I look back on the last decade of Ubicomp research in mental health and use this as an stepping stone for discussing current and future research opportunities. The historical review is based on two recent surveys that I’ve been part of. The first survey presents a review of 45 systems presented over the years and investigate which mental health disease they are designed for, as well as their technical features in terms of sensing, prediction, intervention, and clinical assessment [1]. The second survey investigate whether changes in depressive symptoms can be detected by monitoring the patient’s behaviour using mobile and wearable technology – a core research goal in early research. We reviewed 46 studies, collection more than 17 different features and investigated whether these many studies agree on the relationship between depressive symptoms and patient behaviour collected from mobile and wearable devices. The review shows agreement across studies that some behaviour is strongly correlated to changes in depressive symptoms, while others show no or conflicting correlation [2].

Based on these two surveys, I will discuss current opportunities for research in ubicomp and mental health. In particular I will provide an example of moving from sensing to intervention technology and will present our current technological work in supporting this.

  1. Bardram JE, Matic A. A Decade of Ubiquitous Computing Research in Mental Health. Unpublished manuscript. 2019.
  2. Rohani DA, Faurholt-Jepsen M, Kessing LV, Bardram JE. Correlations Between Objective Behavioral Features Collected From Mobile and Wearable Devices and Depressive Mood Symptoms in Patients With Affective Disorders: Systematic Review. JMIR Mhealth Uhealth 2018;6(8):e165. DOI: 10.2196/mhealth.9691. PMID: 30104184