DataProphet Mobile App

Pocket Square Development assisted DataProphet in deploying their proprietary IP in the form of a mobile application for one of their clients using computer vision technology to identify faults in cast parts. I particularly enjoyed this project as it was the first time I had been involved in the development side of producing a product, and not only the design and management side.

Our Goal

We needed to develop a mobile application for an Android handheld device that allowed the user to take a picture of the part in question, subject that picture to analysis via API and store the result in a database.

We decided to use Android Studio and develop the application in Java at the request of DataProphet. Reason being, they felt Java was a more mature platform on which to build and did not want to rely Kotlin for a relatively simple application that would end up being critical to the normal functioning of the factory.

After numerous discussions expanding upon the system structure and data flow, I began designing a potential UI in Figma.

Designing the UI

My first experiences in development came from the side of a designer. I have a special interest in technical UI and validation, a skill that would be necessary in designing the interface for this particular application. Some initial conditions to consider were the device A Zebra TC8300 Touch, a touch-input handheld computer. This by itself is not a problem, touch interfaces being a standard for quite some time, however users needed to be able to complete a multistep process whilst wearing heavy gloves and (often dusty) eye protection, essential while working with cast parts. This meant any UI elements needed to be bright, clear and large.

The part I love about technical UI is the knock-on effect. The device defined the input method, and the environment constrained the UI, however the constraint then moved in direct contrast to the device. Large, clear icons on a small, low resolution screen would prove to be an interesting challenge that involved validating the essential functions users would need each day, and nesting the rest behind (easily accessible) menus. I thoroughly enjoyed designing a UI where industrial function was the main concern over purely visual design.

Our Process

Being a junior developer at the time, I predominantly followed instructions from Jan, my original Jedi Master. I handled aspects of the frontend implementation where I first learned about best practices, syntax and Git (still not great at those things, I'm a hacker at heart, but I learned about em!), as well as managing client relations with DataProphet. After several iterations, we also had to retrofit a mild scope-change as the potential for the app grew in the eyes of the clients.

Throughout, Jan would assign tasks through JIRA and I would take a ticket. After pushing my code, Jan would review and reflect, merging functional code and providing feedback on issues. Rinse and repeat until complete.

After Thoughts

Working with DataProphet was a dream for a new company such as ours. They were clear in their needs, understanding of the development process and quick to assist in any way they could throughout. If you want to get to know a company, ask their contractors, and we think DataProphet was awesome. This project was a roller coaster from my side as a new developer and I can only thank everyone involved for taking the time to teach me and having the patience for me to learn.