-
Notifications
You must be signed in to change notification settings - Fork 0
Home
Skamba edited this page Mar 28, 2026
·
4 revisions
Reviewed against commit 92a6582.
VibeSensor is an offline-first vibration diagnostic kit for cars. A Raspberry Pi creates its own Wi-Fi hotspot, wireless sensor nodes stream accelerometer data to it, and the browser dashboard turns a road test into live graphs, saved run history, and PDF reports. You do not need internet access and you do not need to install a phone app.
- Get the hardware
- Initial install
- Configure VibeSensor
- Run a diagnostic drive
- Troubleshooting and maintenance
- Assemble a Pi plus one or more wireless sensor nodes.
- Download the latest Pi image from the GitHub repo and write it to a microSD card.
- Boot the Pi, then connect your phone, tablet, or laptop to the
VibeSensorhotspot. - Open
http://10.4.0.1and use Settings -> ESP Flash to program each sensor node from the Pi over USB. - Add the car you are testing, assign sensor locations, and choose a speed source.
- Use the Live view to confirm the sensors are streaming and the graph has real movement.
- Record a drive, then open History to review the diagnosis and export a PDF report.

Use Live while mounting sensors, checking signal quality, and recording a test drive.

Use History after a run to compare sessions, review the strongest findings, and export customer-facing reports.

Use Settings to choose the active car, tune analysis inputs when needed, and select a speed source.
- Minimum bench kit: 1 Raspberry Pi, 1 sensor node, and 1 browser device.
- Typical road-test kit: 1 Raspberry Pi plus multiple sensor nodes so you can compare corners of the car during the same run.
- Expanded diagnostic kit: add an engine-bay sensor and a GPS receiver if you want automatic speed instead of manual speed entry.
- Top-level README:
README.md - Hardware BOM:
hardware/README.md - Pi image build/install reference:
infra/pi-image/pi-gen/README.md - Server config reference:
docs/configuration_reference.md - Field troubleshooting runbook:
docs/operational-runbooks.md