Designing a Data Analytics Platform for Manufacturing

Designing a data acquisition and analytics platform to improve process efficiency.

Role UX Design, Visual Design, Research

Timeline 3 months

Team Architect, PM, TPM, Engineering team

Tools Sketch

Oden Technologies provided realtime manufacturing process data to its customers, but these customers weren’t realizing the value they’d signed up for. Why?

I visited factories, benchmarked other tools, interviewed key roles, and talked with our sales team. From this, I built a map of the problem space (starting at the stated desired outcome), working backwards to identify the most significant blockers and opportunities through discussion with the internal team.

Armed with hypotheses, I began exploring ways of surfacing data in the platform, making insights proactive, and adding tools to classify and find machine data.

I learned that the data lacked context to make it actionable. We’d need to learn what the data meant. This involved a pivot away from trying to figure out how to “package”current data, toward getting the data classified.

We explored automated ways to classify the data, tooling for our prior target user to classify the data post-hoc, and tools for people closest to the machine to classify the data in real-time.

Classified data, available in realtime, through a React native tablet application deployed at the factory line. The ability for engineers to spot good and bad runs in a realtime TV monitoring view, allowing them to rapidly triage. Stored contextualized data by line, product, run, operator, for deep analysis.

This contextualized data was the foundation of Oden’s AI efforts, making for straightforward analysis and performance suggestions.