Reliant on Hardware: Software Testing

Problem

The company I worked for at the time made software for live broadcasting on phones. The customer was interested in seeing a complete version of our product, our software implemented on a portable device. At this stage, we had the software product but had not created hardware. Typically what I would do is build out the hardware chip to enable our software scenarios; this process would take 1-2 years for completion but our customer wanted it now,

Actions taken

We took the Minimum Viable Product (MVP) approach. We built a phone the size of a desktop; the hardware mapped exactly to a that of a phone but on a much bigger scale. We took a GoPro-like camera and put a capture card on the PC, built out a coding/gaming desktop and wrote a small app that behaves as the camera app on a phone. We were able to make a second iteration of our MVP at the size of a laptop within the time allotted from our customer. Our customer was then able to take the product outside and test it on the streets to get a good feeling of the quality of the software product.

Lessons learned

The MVP approach worked out well for us in this scenario. You have to know where you can cut time and resources, and for us in this scenario, it was not necessary to spend hundreds of employee hours creating a perfect phone to test our product. "If I had to do it again I would do most of it the same way. Carrying around the laptop for testing was not a problem so I would definitely do that again." "I would, however, spend more time building out the UX/ UI of the script on our prototype so it would more closely resemble the typical user experience from a software prospective."