Designing a CAN Network Monitoring Dashboard for Automotive Engineers
Designing a CAN Network Monitoring Dashboard for Automotive Engineers
Designing a CAN Network Monitoring Dashboard for Automotive Engineers
Designing a CAN Network Monitoring Dashboard for Automotive Engineers
Designing a CAN Network Monitoring Dashboard for Automotive Engineers
Designing a CAN Network Monitoring Dashboard for Automotive Engineers
Project overview
Project overview
Modern vehicles generate thousands of CAN messages every second. For developers and system testers, that means tons of data but little visibility.
When I joined the Oshkosh team, our goal was clear — turn raw CAN data into something people can actually see and act on. So we built a real-time, visual dashboard that helps both developers and testers monitor system performance without drowning in data.
Modern vehicles generate thousands of CAN messages every second. For developers and system testers, that means tons of data but little visibility.
When I joined the Oshkosh team, our goal was clear — turn raw CAN data into something people can actually see and act on. So we built a real-time, visual dashboard that helps both developers and testers monitor system performance without drowning in data.
Team
Team
Rohan Roy, Bailey Kau, Jenil Makwana, Soumya Kataria, Yilin Chen
Rohan Roy, Bailey Kau, Jenil Makwana, Soumya Kataria, Yilin Chen
Duration
Duration
4 months
4 months
My Role
My Role
UX Designer · Researcher
UX Designer · Researcher
Team
Rohan Roy, Bailey Kau, Jenil Makwana, Soumya Kataria, Yilin Chen
Duration
4 months
My Role
UX Designer · Researcher
Challenge
Challenge
How might we make live CAN data easier to understand for both developers and testers?
How might we make live CAN data easier to understand for both developers and testers?
When I joined the project, the previous team had already designed an initial version of the dashboard. The overall structure was solid — but as the system grew, the data became more complex, and users found it hard to make sense of everything happening in real time.
The main challenge I took on was identifying the most urgent user needs and refining the core features to reflect real workflows. While the initial design was well-structured, we needed deeper insight from end users to validate assumptions, prioritize improvements, and clarify which directions to pursue.
When I joined the project, the previous team had already designed an initial version of the dashboard. The overall structure was solid — but as the system grew, the data became more complex, and users found it hard to make sense of everything happening in real time.
The main challenge I took on was identifying the most urgent user needs and refining the core features to reflect real workflows. While the initial design was well-structured, we needed deeper insight from end users to validate assumptions, prioritize improvements, and clarify which directions to pursue.
Approach
Approach
Listening to the Users
Listening to the Users
To dig deeper and better understand how each role interacts with CAN data, I conducted 15+ short interviews and workflow observations.
The goal wasn’t just to gather feedback on the interface, but to learn how people worked: what data they looked at first, how they debugged issues, what slowed them down, and where the current tools fell short. Through these interviews, I identified two key user groups — one focused on stable, long-term monitoring and the other more concerned with real-time alerts and debugging.
To dig deeper and better understand how each role interacts with CAN data, I conducted 15+ short interviews and workflow observations.
The goal wasn’t just to gather feedback on the interface, but to learn how people worked: what data they looked at first, how they debugged issues, what slowed them down, and where the current tools fell short. Through these interviews, I identified two key user groups — one focused on stable, long-term monitoring and the other more concerned with real-time alerts and debugging.
Through this series of research, I identified the following pain points and opportunities:
Through this series of research, I identified the following pain points and opportunities:
A scalable and modular layout improves usability
Visual context significantly enhances comprehension
Engineers want smarter, not more data
Users need both real-time and retrospective views
A scalable and modular layout improves usability
Visual context significantly enhances comprehension
Engineers want smarter, not more data
Users need both real-time and retrospective views
Existing Dashboard
Existing Dashboard
Difficult to interpret raw CAN logs
Difficult to interpret raw CAN logs
Static updates only
Static updates only
Limited filtering
Limited filtering
Manual, slow process
Manual, slow process
Minimal sharing features
Minimal sharing features
Developer Needs
Developer Needs
Detailed signal visualization
Detailed signal visualization
Stream-based updates
Stream-based updates
Adjustable filters and data channels
Adjustable filters and data channels
Highlight anomalies in real time
Highlight anomalies in real time
Shareable debugging sessions
Shareable debugging sessions
System Tester Needs
System Tester Needs
Simplified overview
Simplified overview
Live dashboard updates
Live dashboard updates
Role-based view
Role-based view
Visual alerts and logs
Visual alerts and logs
Easy report export
Easy report export
Existing Dashboard
Difficult to interpret raw CAN logs
Static updates only
Limited filtering
Manual, slow process
Minimal sharing features
Developer Needs
Detailed signal visualization
Stream-based updates
Adjustable filters and data channels
Highlight anomalies in real time
Shareable debugging sessions
System Tester Needs
Simplified overview
Live dashboard updates
Role-based view
Visual alerts and logs
Easy report export
Solutions
Solutions
A Streamlined Design System and Reimagined Information Architecture
A Streamlined Design System and Reimagined Information Architecture
To address the above issues, I added around three targeted improvements:
To address the above issues, I added around three targeted improvements:
Real time dashboard
Real time dashboard
The dashboard visualizes thousands of CAN signals in real time, converting raw message streams into live data curves.
Developers can zoom into signal details, while testers monitor system performance at a glance.
This real-time flow turns complex logs into a clear, intuitive data story.
The dashboard visualizes thousands of CAN signals in real time, converting raw message streams into live data curves.
Developers can zoom into signal details, while testers monitor system performance at a glance.
This real-time flow turns complex logs into a clear, intuitive data story.
Developer View
Developer View
During research, I learned that developers were constantly buried in messy CAN logs and static charts, making debugging slow and frustrating. They needed more control and a clearer way to see what was really happening in the data. So I designed this view to give them just that — real-time graphs, sortable signal tables, and built-in error tracking. It turns complex data into something visual and interactive, helping developers spot issues in seconds instead of hours.
During research, I learned that developers were constantly buried in messy CAN logs and static charts, making debugging slow and frustrating. They needed more control and a clearer way to see what was really happening in the data. So I designed this view to give them just that — real-time graphs, sortable signal tables, and built-in error tracking. It turns complex data into something visual and interactive, helping developers spot issues in seconds instead of hours.
Tester View
Tester View
While developers needed detail, testers needed clarity. From the interviews, I found that testers cared less about raw data and more about system behavior — whether everything was running smoothly or not. This view focuses on high-level performance and alerts instead of technical frames. With clear gauges, color-coded statuses, and trend visuals, testers can quickly understand system health, spot warnings, and keep the big picture in sight without digging through data.
While developers needed detail, testers needed clarity. From the interviews, I found that testers cared less about raw data and more about system behavior — whether everything was running smoothly or not. This view focuses on high-level performance and alerts instead of technical frames. With clear gauges, color-coded statuses, and trend visuals, testers can quickly understand system health, spot warnings, and keep the big picture in sight without digging through data.
Reflection
Reflection
Designing for Two Mindsets
Designing for Two Mindsets
Working on this project really showed me how important early research is — especially for something like a dashboard. You can’t just start designing pretty charts; you have to know what users actually need and what information really matters to them. Talking with developers and testers helped me see those differences clearly and design around their priorities.
This capstone course also helped me grow a lot in how I present ideas and work with a team — learning to explain my design choices, listen to feedback, and build something together that truly works for its users.
Working on this project really showed me how important early research is — especially for something like a dashboard. You can’t just start designing pretty charts; you have to know what users actually need and what information really matters to them. Talking with developers and testers helped me see those differences clearly and design around their priorities.
This capstone course also helped me grow a lot in how I present ideas and work with a team — learning to explain my design choices, listen to feedback, and build something together that truly works for its users.
© Yilin Chen 2026. Proudly hand created.
© Yilin Chen 2026. Proudly hand created.
Redesigning for Trust
An Energy Website Revamp That Converts