Smartphones are increasingly powerful, sensor-rich and personal devices. Could they also help us to become happier and healthier? Dr Neal Lathia discusses how we use smartphone sensors to track how people behave, and design systems that use these inferences to help us all become healthier.
Dr Lathia is speaking during the event, Happier and healthier with smartphone data.
CSF: There’s been quite a lot in the media lately about the negative aspects of using smartphones, ie privacy issues. However, can you describe the positive aspects based on the work you do?
NL: I agree that it is increasingly important to question the sensitive nature of the data that smartphones can collect. But my research has always focused on the prospective benefits of carrying engaging devices that can sense and learn – there are so many aspects of our daily lives where knowing more about ourselves and/or our surroundings could provide substantial help. For example, smartphones can help us explore our city and discover social events we may be interested in, capture health symptoms as the occur, discover and report transport disruptions, and guide us through behaviour change interventions, such as smoking-cessation programmes.
CSF: How do smartphones track our moods and happiness?
NL: The Emotion Sense app tracks mood and happiness by acting as an intelligent journal. The app asks you, at random times of the day, to fill out a short survey about how you currently feel. The reasoning behind this well-known research method is that it is easier for you to answer questions about the here and now, rather than remember how you felt yesterday, last week, or last month. In the background, the app also takes snapshots of sensor data from your smartphone. For example, it measures how physically active you are, or how socially active you have been via your phone that day (by counting how many phone calls you have made).
CSF: What good can come from this?
NL: Every small interaction with the app is akin to giving it a piece of a huge puzzle. The app then takes care of putting all the pieces of the puzzle together for you, and it shows you the relationships between the questions you answer and the sensor data that was passively collected. There are over 40,000 people who have downloaded this app since it was released in 2013, and I've received a variety of emails where people tell me of the benefits they experienced – some of them were even beyond what we originally intended the app to do. One user discovered when he feels the most creative; others told me how tracking their mood this way increased their productivity. Another user found that the app helped to manage and predict fluctuations in bipolar symptoms.
CSF: How will our phones be able to improve our happiness based on the data it has collected?
NL: Happiness is not an easy thing to improve – it is ultimately a subjective experience that can vary quite substantially between individuals. But our phones can help us become more aware of our surroundings, behaviours and moods: and becoming aware of something is a critical step in learning to manage, understand and improve it.
CSF: So will are phones become our therapists?
NL: In the near term, I don't think that phones will become our therapists. Instead, they will empower us to become better therapists for ourselves. There is also a growing recognition that this data could help our clinicians understand us better, and have to rely less on patients recalling their symptoms when providing care. For more information, visit Opportunities for Smartphones in Clinical Care: The Future of Mobile Mood Monitoring.
CSF: Has it so far yielded positive results? What have you discovered?
NL: An interesting result that we found relates our daily physical activity to our happiness. There has been a lot of recent research warning us about the health risks of a sedentary life. The Emotion Sense data is showing us that a similar relationship exists with our happiness. I'm sure that there are many more results yet to come.
CSF: Are there any challenges / limitations to this technology?
NL: Collecting this data in a privacy-aware manner remains challenging, as we have to balance between collecting data and potential research benefits. For example, one of the ways that Emotion Sense characterises your environment is by taking a short recording from the microphone. The researchers do not receive any audio recordings – they only receive a measurement of how loud or quiet the environment is around you. Naturally, this has set back what the researchers can and cannot look into, but is an example of a choice we made to favour privacy over research outcomes.
The challenge that interests me the most is about how to make sense of this data and put it to good use; this means moving away from collecting data for research analysis and towards systems that can apply machine learning in the right way to help us change and improve our behaviours. The research on behaviour change and smartphone sensors has historically been in different fields: I believe that the intersection of the two is full of interesting opportunities.
CSF: How do you see this work developing in the future?
NL: The data that Emotion Sense has collected remains incredibly unique. The results that I mentioned above used only two dimensions of the data (physical activity and happiness), and there are many others that the app captures (location, personality, sociability, communication patterns, and so on). I’m sure that there are many years of research yet to come from this dataset alone. Also, other researchers are now using our open source tools to study, for example, how we react to prompts from our phone and how people with depression move. In the future, as wearable sensors become more popular, the potential to collect fine-grained behavioural data will grow. This is going to offer an unparalleled opportunity for researchers to study daily life in a way that was previously inaccessible.
CSF: What are you currently working on?
NL: I have recently left the Computer Laboratory, and am now working out of Accelerate Cambridge at the Judge Business School. I've developed a platform for researchers to conduct smartphone studies (called Easy M) and have recently completed a number of pilots using it with various partners around the world. I’m now using all of that experience to design new systems that can take steps towards the grand challenge discussed above – making sense of the data and putting it to good use. You can follow my journey on twitter (@neal_lathia) or Medium (https://medium.com/@neal_lathia).
Image copyright: Neal Lathia Emotion Sense