According to project partners at Southampton University, the prototype contains removable sensors that monitor heart and skin activity. The data collected is processed via a model to determine the emotional state and the intervention is sent to the wearer via a smartphone app.
M.c. schraefel, a professor in computer science and human performance design from Southampton University helped design the system.
She told The Engineer: ‘There are two types of hunger…one is homeostatic, which means we are eating to get our bodies back into balance: basically, we’re hungry because we need fuel and nutrients. The other is referred to as hedonic, its eating that nothing to do with physiological requirements at the time.
‘There’s an area of work…that talks of emotional eating, where we tend to eat at times of stress. Some people will eat when they get into that…state as a comfort food response - and some people will not eat at all - so there’s a complexity in terms of responses to the same kinds of stimulation.’
The aim of the project, published in a study entitled ‘Food and Mood: Just-in-Time Support for Emotional Eating,’ was to develop a system that could distinguish between states and make interventions when appropriate.
Schraefel said: ‘the question is: if you’re feeling triggered and there’s an emotional response and its physiologically based…then is that signal, the physiological change, distinct enough for us to be able to say, “this is the type of emotional signal - or signature - for emotional eating versus the physiological signature for [being happy]” Are they sufficiently distinct when…a machine’s looking at them to be able to tell the difference?’
To build their model the team, including researchers from Microsoft Research and the University of Rochester, US, had users log their emotions and what they had eaten every hour via the app.
‘The accuracy of the model was about 75 per cent in terms of detecting an appropriate kind of emotional state,’ said schraefel.
The smart bra then added physical data to the emotions so they can be detected without prompting the user to log every hour.
According to the paper, the system used GRASP (Generic Remote Access Sensing Platform), a custom-build real-time device made up of a sensor board, firmware, software libraries, and an API. GRASP also incorporated an MSP-430 microprocessor and is powered by a 3.7V Lithium-Ion polymer battery. In use, GRASP can sample up to eight bio-signal channels simultaneously.
The GRASP boards in the study were configured to capture heart rate and respiration with an electrocardiogram (EKG) sensor; skin conductance with an electrodermal activity (EDA) sensor; and movement with a three-axis accelerometer and a two-axis gyroscope.
Schraefel said the batteries struggled with the amount of data that needed processing in real-time. She said similar systems might record signals for a period of time, log them, then burst send them to a device to save on the battery power.
‘In this case we weren’t doing that…a steady stream of information is being passed to the circuit board, which has to be on to process this information, and is then sending regular updates by wi-fi to phones that are on the person,’ she said. ‘We’re trying to save some energy but there are probably efficiency gains to be made.’
She added: ‘To really be able to help people…it would need to be tried with a much larger population. The challenge of doing that is…trying to build an apparatus that is robust enough and cheap enough to produce to be able to distribute it to a [number] of people.’
According to the paper, there are ‘several future research directions that are essential to building an integrated system for just-in-time support.
Schraefel stressed that work in this area includes investigating interventions for men with similar eating anomalies.
UK productivity hindered by digital skills deficit – report
This is a bit of a nebulous subject. There are several sub-disciplines of 'digital skills' which all need different approaches. ...