Emily Phan

Google's Project Tango: Dora

A 3-D solution to crime scene documentation using the Google Project Tango.


The Story

This was a 2 month internship project at TASER international. I collaborated with three engineers interns who are Alex Choi, Micah Rosales, and Minh Pham to create a solution for crime scene investigations. Our task was to use Google's Tango to improve crime scene documentation. 




A crime scene documentation solution that uses 3D technology. Uses Google's Tango, an android tablet to scan and create a 3D model of a room. Able to interact with models on a tablet and web application. 




The design process began with researching about the crime scene investigation procedures to identify strengths and weaknesses to find opportunities where design could intervene. Unfortunately, our team was unable to talk to investigators; therefore we relied heavily on guides. Also, we researched existing 3D technology that are being used in crime scene investigations. 


Problem Space

Crime scene documentation is inefficient and ineffective. It can be a complex process and currents solutions are limited.  


System Modeling


I devised conceptual models to identify core components and visualize the flow of the interactions. By doing this I was able to understand what I needed to design and it helped me communicate with the engineers to see what had to and could be built in our time frame. 


Design in a Spectrum


Designing variations isn't just helpful for me as a designer to explore ideas; however it allows me to give options to engineers. By doing this we were able to bounce ideas and inspire each other to create a meaningful product. Even though towards the end we decided it would be the best to simplify the interactions cause of time constraints; however we saw the potential of our product growing. 


Technical Challenges

There were a lot technical issues that engineers encountered and it greatly affected the design process. There were many inaccuracies and unknowns that led to many design changes. For example, the inaccuracy of recording the room to get the 3D model and marking the evidence on the model. 


Design Challenges

The three biggest design challenges were teaching a person how to scan a room, letting the user know when to stop scanning the room, and figuring the process to document a piece of evidence. I asked myself many times can this be solved by changing the flow of the interactions or adding graphical elements? Through many iterations of wireframing, prototyping, and testing we were able to overcome these obstacles. 


Document Evidence

An interaction design problem I encountered was figuring out how the user would interact with the 3D model. Therefore I started to draw wireframes; however I still couldn't visualize it. Therefore I mocked up a paper prototype and it really helped me decide and communicate my design. From there I made multiple screen state flows to find the smoothest interaction. 


Material Design
Gone Wrong

A lot of technical issues occurred that changed how we envisioned the product. For example, we had come to an agreement that we would use material design since it is a Google product. However, we discovered that the tablet was running on KitKat instead of Lollipop. We were able to implement a majority of Google material design however, the user interface is a hybrid of material design and not material design. 


One Step vs
Two Step

Initially, I had separated two tasks which were creating a marker and viewing the evidence list; however I saw that both actions were very similar to each other. Therefore instead of separating the two actions, I decided to combine them in one button that would allow the user to view their evidence list as they create new markers. 


Product Walkthrough

Below is a walkthrough of the tablet application. 


Final Thoughts

I learned so much from this project such as articulating my design ideas and standing up for myself as a designer to learning how to use google material design. I learned how to compromise and be okay that everything I envision does not have to be fully implemented.