Problem Discovery 

Why make this tool?

Prior to this tool, auto companies would use crash dummies to develop and test vehicle safety features.
‍
A crash dummy mimics the body type of a 30-year-old, 6-foot-tall 180-lb male. That leaves the rest of us in vehicles that aren't designed for us, which makes us (and the rest of the population) morbidly vulnerable in auto accidents.
‍
Due to its location in the University of Michigan hospital and Michigan's entwined history with Big Auto, this research department had been collecting bodily data on auto accident patients for decades. They had a generous data set with predictive measures for crash victims--much more accurate for vehicle design than using crash dummies.
‍
They just needed a way to make that data available to the auto companies.

Research

User Interviews
‍I interviewed 3-5 crash researchers and asked questions like:

How often do the auto companies approach you for data?

How do they contact you when they need this data? (phone, email, in-person, etc)

What type of information is the most commonly requested?

This allowed me to gather qualitative and quantitative research.
‍
Customer Visits:
‍Once a month, the research department would host a gathering called Case Review, where representatives from various auto companies (Toyota, GM, Chrysler, Mercedes-Benz) would visit us and watch a presentation on outlier vehicle crashes that occurred that month.

After the presentation, I talked to some of the reps one-on-one to find out what data they were looking for from the department and why.
‍
Heatmap
‍I used a heatmap plugin to find the hotspots of where users were clicking on our tool. This helped me refine the design through each iteration.

Design Process

From this research, I learned that both the auto companies and researchers wanted to give us the age and gender of a person in a vehicle, input that into the data, and in return be able to compare male and female data sets side-by-side, their projected weights, and the percentile of where they fall in the body index data set.
‍
This problem sounded complex, but when I abstracted it to [person] wants to [do a thing] so they can [goal], it became a lot clearer. The research told me how to fill in the blanks: [Auto researchers] wanted to [give specific data to us] so they could [get other specific data and use it to design safer vehicles.] Or simplified even more: [Researchers] wanted to [give us numbers] so they could [get different numbers].
‍
And what tool do we use where we put in numbers and it spits out different numbers? Sound like a calculator to you? Sounds like a calculator to me, too.
So, I began designing the interface for this fancy calculator with pencil sketches.

The Conclusion

With that calculator, auto companies were able to easily look up measurements and percentiles for various body types, rather than just using the standard one-size-fits-all crash dummy.
‍
Through research and user testing, I learned that the calculator was the priority--all those graphs and scan pictures and other data was just frosting on the cake.

To this day, auto researchers still use this tool, so I'd call this project a success.