Not only are the products we use becoming ever smarter and generating mountains of data, they are increasingly connected to the internet and each other and forming intelligent ecosystems. Often this data is used retrospectively to evaluate what people think of a product. That's odd, think TU/e doctoral candidates Janne van Kollenburg and Sander Bogers. “We are saying the reverse: data has become such an important part of the design that you should base the design process on the data being gathered during product trials.”
Joint PhD conferral
That the researchers would write a joint thesis was not the original plan. More than four years ago they each embarked on their own research question, but it soon became apparent that their work was complementary and, sometime later, that their research stories could not in fact be told separately. Now they have co-authored a hefty piece of work. Just in time; TU/e recently decided to abolish doctoral duos.
But it is the new design method, stresses Bogers, that makes this work special. “A designer will often start the thinking process on paper - a list of sensors and what they need to do, then an initial sketch. But what actually happens in practice is that the value of data is often overestimated, or conversely underestimated, and that means unique, valuable patterns are missed. Data-enabled design, the method we are developing, regards data as an essential part of the design itself and of the design process.” “In our process we use the uniqueness of the individual instead of a common factor, and we take a new and different view of the data,” explains Van Kollenburg further. “In order to get inspired and to adjust the design along the way, rather than to carry out an evaluation. In this way, a product can grow and evolve over time into a self-learning system that benefits from the input of a designer every now and then.”
Smart Baby Bottle
In cooperation with Philips, Van Kollenburg and Bogers further elaborated their method using various examples in which data lay at the heart of the design.
One such example was the Smart Baby Bottle, whose commercial launch is imminent, Van Kollenburg tells us enthusiastically. “This removable cover, which fits a baby bottle, keeps an accurate record of such data as when and for how long a baby drinks and the temperature of the fluid consumed. A related app can also keep track of breastfeeding and milk expression sessions. We have seen, for example, differences in milk temperature and drinking behavior when a baby is being fed by the mother or the father and we can give tips based on the data. A movement sensor in the bottle reveals when bottles are taken in the car, and a subsequent feed is given in an entirely different setting rather than at home in the nursery. Such personal examples help shape a design. How can you optimize feeds and offer parents more clarity? How can you help them reduce the number of feeds or to sleep a little better at night?” Bogers, with a wry smile: “Parents are willing to try anything to get a little more sleep.”
Even an invasion of their privacy? “Fairly early on in the test period, parents became very interested in the data that was being collected and started to join us in wondering how a design like this could help them and what kinds of issues needed to be taken into account, such as privacy. For us, that marked the start of a follow-up project in which we made the system experiential. This places any issue of privacy in a different context; parents receive something valuable in return, and that adds a personal dimension to the discussion about privacy. We have learned that which data is relevant is something you need to consider carefully, as is how you then reproduce and represent that data. Invest in creating a dialogue between the user and the design, and be sure to give the user plenty of scope to speak openly.”
Their design method, which is both scientific and practical, is being received in various quarters - “across the board, design strategists from Coca-Cola through to Schiphol are enthusiastic” - as refreshing. In the meantime, the work done by Van Kollenburg and Bogers has already prompted a new Data-enabled Design Team at Philips.
In addition, they will soon start teaching a Master's course in order to train a new generation of ID people who are not afraid to see data as a new design material. For the time being they themselves will be staying on at Philips, working in the new team; there are plenty of plans waiting to be worked up. But first of all, it's party time. And after four years of collaboration, they are more than happy to party together.