PicoDash Telemetry System
Replacing spreadsheet chaos with a real-time sensor dashboard — giving test engineers control over 70+ sensors during hypersonic launch tests.
Test day was a scramble
Longshot's accelerator uses dozens of sensors — pressure transducers, photodiodes, induction coils — to capture data during each test fire. Before PicoDash, managing these sensors meant spreadsheets, terminal commands, and a lot of manual coordination.
On test day, engineers needed to arm sensors, verify connections, trigger data capture, and manage firmware updates — often under time pressure. The existing workflow was fragile: one wrong command could mean missing critical data from a test that took weeks to prepare.
"We'd spend the morning of a test just getting sensors into the right state. It was stressful, error-prone, and took attention away from the actual engineering."
One screen to control everything
I built PicoDash as a unified control interface — a single dashboard where engineers could see every sensor's status and perform bulk operations without touching the command line.
Core Features
- Real-time sensor status (armed/disarmed, connected, last transmit)
- Bulk selection and operations
- One-click arm/disarm for selected sensors
- Firmware upload directly from browser
- Customizable column views per user
Design Decisions
- WebSocket connection for true real-time updates
- Color-coded status (green = ready, red = needs attention)
- Persistent column preferences
- "Classic UI" toggle for engineers who preferred the old interface
From chaos to clarity
Test prep went from hours to minutes
PicoDash became the default tool for test operations. Engineers could arm all sensors with a single click, immediately see which ones weren't responding, and fix issues before they became test-day emergencies.
What I learned
The most important feature wasn't any single button — it was visibility. When engineers could see all 70+ sensors in one view, with clear status indicators, they could make decisions faster. The cognitive load dropped dramatically.
I also learned the value of keeping the old interface around. Some engineers had built muscle memory around the classic UI. Rather than forcing migration, I added a toggle. Adoption happened naturally as people saw the benefits.