Sense
Acquire the race car’s existing brake-pressure signal without disrupting the vehicle’s electronics.
Working end-to-end prototype · Track tested
Independent engineering project · 2024—Present
Brake BuddyTurning braking intuition into measurable performance.
A physical-AI driver-coaching system that transforms brake pressure, GPS position, and session history into real-time feedback and personalized post-session analysis.
Embedded hardwareGNSS contextHuman-machine feedbackTelemetry analytics
One system, built across the full stack.
01Vehicle-integrated hardware
02Real-time embedded firmware
03BLE mobile configuration
04Automated analysis + coaching
The engineering problem
Drivers are asked to hit a number they cannot feel.
Brake pressure is abstract. A coach can prescribe a target, but the driver has no direct reference for translating that number into pedal force while managing a corner at speed.
Brake Buddy closes that loop. It observes the car, understands where it is, gives the driver immediate visual feedback, and preserves the session for analysis after the car returns to the paddock.
Understand
Combine pressure with GNSS position to identify the active track, lap, corner, and braking zone.
Respond
Translate live data into intuitive LED feedback against a corner-specific pressure target.
Learn
Turn logged sessions into reports, history, and driver-specific AI-assisted coaching.
Brake pressureGNSS positionDriver history
CONTEXT ENGINEBrake BuddySense · Localize · Compare
Live LED feedbackCorner targetsPost-session coaching
System architecture
Designed as a complete coaching loop—not a standalone light bar.
A fused vehicle interface feeds a custom ESP32-S3 system. GNSS adds track context, BLE exposes configuration, and the SD-card workflow carries session data into the local analysis platform.
Real-time driver feedback
Information at the moment it can change the next corner.
The dashboard-mounted DotStar strip converts live pressure into a peripheral visual reference. The driver sees buildup, recognizes the target window, and receives warnings without reading a conventional display.
- Pressure buildupProgressive spatial feedback below target
- Target achievedGreen flashing inside the calibrated window
- Excess pressureImmediate red warning beyond the target
- No-brake zoneTrack-aware warning in defined regions
An integrated platform
Hardware, mobile control, and analysis—built as one product.
Each layer was developed around the same constraint: turn complex telemetry into feedback a driver and coach can use immediately.
01 / HARDWARE
Vehicle-ready embedded sensing
An ESP32-S3-based unit integrates signal conditioning, regulated automotive power, GNSS, BLE, persistent configuration, and local data logging inside a track-tested installation.
- Existing 0–5 V pressure sensor input
- SparkFun NEO-M9N GNSS
- OpenLog microSD logging
- Fused 12 V vehicle integration
02 / MOBILE
Configuration in the paddock
The iPhone app communicates with the embedded unit over BLE, giving the coach control over track selection, global and corner targets, driver profiles, LED behavior, and a live pressure demonstration.
03 / DATA MANAGER
Drop in the SD card. Keep the sessions that matter.
The local Brake Buddy web app accepts every file from the SD card in one pass. It filters out unusable logs, imports the meaningful on-track sessions, attaches driver and car context, and automatically generates the reports worth reviewing.
- 01Select every SD-card file
- 02Filter for usable on-track data
- 03Create sessions and reports
04 / REPORTS
From an SD card to a coaching conversation
Each usable session becomes an automatic, browser-based report. Brake Buddy detects laps and braking events, maps them to the circuit, and turns raw pressure traces into a review a driver and coach can discuss immediately.
Brake-event detectionTarget performanceTrail-braking analysisConsistency and fatigueTrack mapsHistorical context
05 / AI COACH
A coaching conversation grounded in the driver’s own data.
The AI coach can compare selected sessions, surface repeatable strengths and weaknesses, explain techniques such as trail braking, and translate the report into a focused objective for the next run. Driver history keeps the conversation connected across sessions.
AI coaching is decision support—not a replacement for a human coach or a claim of autonomous instruction.
Driver feedback
What changed when drivers could finally see the brake pedal.
Selected comments from track testing, anonymized for this portfolio.
“I think this is going to be so beneficial to coaching”
“It works great, it is there when I want it, when i don’t need it, it hides well and is not distracting.”
“This helps you get consistent on braking”
“It feels so much easier to quantify brake pressure in color and brightness”
Product filmApprox. 02:00
In production
Coming soon
See the complete system move from car to coaching.
The final two-minute film will cover the hardware, live LED feedback, GPS context, mobile control, reports, and AI coach.
Validation, not just assembly
Built, calibrated, installed, and tested where it matters.
Brake Buddy has progressed from a bench prototype to an end-to-end system used during real track sessions. The current prototype has been evaluated by multiple drivers and coaches.
Vehicle-safe integrationDesigned around the existing race-car sensor and fused power system.
Telemetry calibrationPressure readings validated against the car’s AiM data.
Track-session reliabilitySensing, feedback, positioning, and logging exercised together on track.
Human feedbackDrivers and coaches evaluated the usefulness of the live reference and reports.
Development journey
Each phase expanded the question Brake Buddy could answer.
The system grew through track feedback and engineering constraints—not a predefined feature checklist.
Pressure sensing
Vehicle-safe signal acquisition, conditioning, and calibration against AiM telemetry.
Live feedback
Progressive LED behavior, target windows, over-pressure alerts, and reliable SD logging.
Track context
GNSS track identification, lap detection, brake zones, and corner-specific targets.
Mobile control
A BLE iPhone workflow for track setup, pressure profiles, LED settings, and live demo.
Data + AI
Local session management, automated reports, driver history, and session-aware coaching.
Engineering depth
The hard work lived between the disciplines.
The project required mechanical judgment, automotive electrical integration, embedded software, geospatial logic, data analysis, and human-machine interaction to work as one dependable system.
- 01Non-destructive vehicle integration
- 02Automotive power and electrical noise
- 03Pressure calibration against telemetry
- 04Reliable embedded logging
- 05GNSS corner and lap detection
- 06Persistent BLE configuration
- 07Actionable telemetry metrics
- 08Driver-specific historical context
Engineering ownership
Conceived, integrated, and validated by Alan Piao.
I independently developed Brake Buddy from the original driver problem through vehicle integration, embedded firmware, track mapping, mobile workflow, reporting, and product direction. AI development tools were used as engineering assistants throughout the process.
Product conceptSystem requirementsHardware selectionVehicle integrationFirmware developmentSensor calibrationTrack mappingData analysisMobile workflowAI architectureTrack testingProduct direction
About the builder
Wuxuan “Alan” Piao
Mechanical engineer · Motorsport engineer · Physical-AI builder
Purdue University Mechanical Engineering graduate and motorsport engineer at Team Stradale. My work combines race-car data analysis, mechanical design, embedded systems, software automation, and driver-performance engineering.
I am interested in graduate study at the intersection of robotics, intelligent vehicles, cyber-physical systems, and human-machine interaction.
Photography is my way of staying curious about light, motion, and people. See my photography at alanpiao.com