Kenworth — Test Reporting & Diagnostic Dashboard

During my Kenworth internship, I developed a Power BI dashboard connected to a SQL database to automate test reporting and help engineers investigate electrical and software test failures across PACCAR plants. The dashboard replaced a manual Excel reporting process and brought test status, trends, and diagnostic information into one place.

Each truck underwent more than 40 electrical and software tests. Determining whether testing requirements had been met across locations involved querying the database, manually interpreting pass/fail rules, and updating an Excel document.

Beyond automating that process, the team needed better visibility into failures. With more than 11 million rows of data per software version, the challenge was turning a large volume of results into information people could use to diagnose problems.

SQL and Power BI were both new to me. I learned to query the database, translate the existing test-status rules into automated logic, and build accessible views.

I organized meetings with the people using the software and diagnosing failures. I asked how they investigated a problematic test and used their answers to shape comparisons of first-attempt passes with successful retests, pass/fail rates, and recent test histories.

I released the first working version halfway through my internship so users could apply it to their daily work. Their feedback revealed opportunities that had not emerged during initial requirements discussions.

Navigation that felt straightforward to me was not always clear to other users. I refined the organization and presentation to make relevant information easier to find.

I expanded plant-by-plant comparisons to help users identify tests that performed differently across locations and investigate those differences together. Additional views surfaced the ten most recent passing and failing results, with test and chassis context to support troubleshooting.

The dashboard replaced the manual Excel reporting workflow, reducing the effort needed to compile and interpret test status. Users could access targeted comparisons and test histories more quickly, helping them investigate failures and support quality improvements.

The value of data depends on how clearly it supports a decision. Effective dashboards require accurate calculations, intuitive navigation, meaningful comparisons, and an understanding of how people investigate problems.

Releasing a working version early was especially valuable. Once users could interact with it, they identified new needs and helped me refine the dashboard into a more useful engineering tool.