Design with Data to Disrupt:
Fusing Design Thinking and System Engineering to Enable Efficient
Disruptive Data-Science!
An old story!
Long long back there was a king who wanted to have a competition to award the strongest subject in the kingdom. It was decided that the men wanting to compete for the title would be given one hour each and a spade to dig a trench. The one who could dig the longest would get the prize! After a lot of preliminary searches, two finalists were chosen for the main competition. With the gun-shot, one man went to dig immediately. The other man took his time to check the spade, polished the handle for a better grip and used a stone to sharpen it. Even though our second competitor started late, he won the competition!
Disrupt, Data and Design-thinking
In spite of the extremely rosy predictions made by consultancy firms there are many major hurdles when it comes to applying data-science and AI to make money!
On one hand, for an established conventional industry it might take a lot of e ort to explain to the stakeholders the need of data-engineering and AI. Many such industries have a sound customer base, are financially stable and do not want to explore unless they have to. Mostly, we go with an application of data-science which we believe will make a difference and it is almost impossible to convince them! On the other end of the spectrum are initiatives we, the data-science entrepreneurs, take in trying to make a new product leveraging on AI. We try to disrupt in both the above cases and we face serious resistances and challenges.
In this blog, we shall discuss a methodology which we have found to be effective where we fuse Design Thinking and System Engineering. The power of Design Thinking (DT) lies in enabling us to gather the most pertinent requirements. It helps us to know the “right thing to design”. Once we have the knowledge of the exact requirements, we use the System Engineering (SE) framework to “design it right”!
It can also be noted that DT is not something which can be done once and done away with! The DT cycle (as will be discussed soon) can be repeated at any phase of the project (if needed). It is mostly pertinent to do it if the performance of the system goes below expectation and intervention is needed.
Design Thinking: Getting the Right Requirements
Of late, Design Thinking has been emerging as a powerful tool to be used by technocrats and entrepreneurs (dschool.stanford.edu). The usual DT cycle involves four major phases, viz. Empathize, Ideate & Define, Prototype and Test.
Figure 1 shows the DT flow which we have fine-tuned for the domain of data-science and AI. The first step is always to empathize with the end-users and to endeavor to understand their pain-points. For example in one of the projects we were exploring how can the shared-taxi services in Cape Town be made safer. After a few rounds of initial informal discussions with the daily commuters and some of the taxi drivers one of the pain-point discovered was that some taxis speed in lonely roads of not-so-safe suburbs to avoid taxi-jacking. Hence, one solution explored was to use a simple interactive mobile application to suggest the driver the route with more number of vehicles (so that they avoid lonely routes). In this phase in addition to exploring the pain-points, we also need to empathize with a possible challenge the end-users might have. For example, in some countries, a mobile phone application based solution may not work because of the lack of inexpensive mobile-data.
In the next stage, we need to ideate as a team, unpack the interview results and try to converge on defining a set of initial needs. From a data-science perspective, this also means that we decide on what data to use and how can such data be collected.
