Project · IoT / Embedded Systems
Micromobility Vehicle Electronics System
At DeltaX Automotive, a micromobility startup focused on last-mile connectivity solutions, I worked as the IoT Engineer Intern on their flagship product 'KIX' – an electric kick scooter designed for 5-10 km urban commutes. My role encompassed the complete electronics system development from embedded software architecture to hardware manufacturing.
The project involved designing a sophisticated authorization system featuring fingerprint-based keyless entry, electronic motor locking, and a swappable battery mechanism with automated door control. I developed the Blynk IoT mobile app for remote vehicle management, implemented CAN bus communication with the VESC motor controller for real-time telemetry, and designed both charging and discharging circuits for the dual-battery system.
A significant challenge arose mid-project when the global chip shortage made Raspberry Pi Zero unavailable worldwide. I led the pivot to ESP32 architecture, porting the entire codebase from Python to C/C++ while maintaining all functionality. Through systematic range testing and motor controller tuning, we achieved the target 50 km range. The project culminated in a working MVP demonstrated to potential investors, showcasing the complete authorization flow from fingerprint scan to motor unlock.
Problem Statement
- Need: DeltaX Automotive required a complete electronics system for their KIX electric scooter, including keyless authorization, motor control, and battery management for a target 50 km range
- Prior state: Existing prototype had basic electronics with reliability issues — relay failures, chassis voltage leakage, and unorganized wire harnesses prone to failure
- Challenge: Mid-project global chip shortage eliminated Raspberry Pi Zero availability, requiring a complete platform pivot without delaying the MVP timeline
- Open question: Could we design a modular, reliable electronics architecture that would enable rapid prototyping while meeting investor demonstration requirements?
Methodology
The project evolved through three major iterations, each addressing discovered issues and improving system reliability. Click a stage to jump there.
Iteration 1: Raspberry Pi Integration
Started by understanding the existing codebase and integrating a Raspberry Pi Zero with the pre-existing daughter board. Added RPM sensor functionality using an ADC for analog-to-digital conversion. Implemented hardware button for fingerprint enrollment and developed Bluetooth command protocol between two Raspberry Pi units for remote control testing.
Issues Critical issues identified
- Relay activation failure: 3.3V GPIO insufficient to trigger 5V relay coils
- Solenoid failure: Caused 14V chassis leakage creating safety hazard
- Wire harness: Solid-core wires with improper lengths causing connection failures
Iteration 2: PCB Design & Chip Shortage Pivot
Designed custom PCB with SMD components for a more robust system. During testing, a faulty component damaged the CPU, coinciding with the global chip shortage that made Raspberry Pi Zero unavailable until end of 2022. Conducted extensive research on alternative microcontrollers, evaluating supply chain reliability and automotive-grade options before selecting ESP32 as the new platform.
Key Decision ESP32 selection rationale
- Supply chain: Widely available despite global shortage
- WiFi/BLE built-in: Enabled IoT app connectivity without additional modules
- GPIO availability: Sufficient pins for all sensors and peripherals
- Community support: Extensive libraries for CAN bus, sensors, and Blynk IoT
Iteration 3: ESP32 Implementation & MVP
Ported entire codebase from Python to C/C++ for ESP32 compatibility. Implemented complete feature set: fingerprint authorization (enroll, authenticate, delete), relay-based motor locking, Hall effect side stand detection, and LED-based HMI. Manufactured final protoboard with JST connectors for modularity. Added solid-state relay to prevent back-current damage discovered during range testing.
Blynk IoT App Development
Designed and implemented mobile app using Blynk IoT platform to replace hardware buttons with software controls. App features include Enroll button for new fingerprint registration, Battery Door control for swappable battery access, and real-time HV/LV battery voltage monitoring.
Wire Harness Design
Evolved harness design across three iterations to eliminate failure points from solid-core wires and improper lengths. Final harness used silicone wire with measured lengths (using thread for routing estimation), JST connectors for modularity, tape bundling for organization, and built-in slack for maintenance access and future expansion.
Charging & Discharging Circuits
Designed custom discharging circuit using series-connected 12V CPU fans to bring HV battery within ESC voltage tolerance (preventing over-voltage protection trips). Created LV charging circuit to charge the 4S Li-ion floor-mounted battery by stepping down 60V HV supply to 30V input for the charging board.
Sensor Integration
Evaluated multiple sensor options for side stand detection: Z-stop switches, reed switches, and Hall effect modules. Selected Hall effect sensor for reliability and non-contact operation. Coordinated with mechanical team on sensor mounting to ensure protection from moving parts and proper isolation from chassis voltage.
Motor Controller Communication
Attempted UART communication with VESC motor controller but encountered library compatibility issues. Successfully implemented CAN bus protocol, establishing reliable telemetry extraction including input voltage, MOSFET temperature, duty cycle, motor temperature, current draw, and tachometer readings. Developed data parsing to convert hexadecimal frames to human-readable values.
Test Setup
System validated through multiple range tests and authorization flow demonstrations with potential investors and stakeholders.
Method Testing methodology
- Range testing: Multiple full-discharge runs with VESC parameter tuning to optimize efficiency
- E-brake testing: Iterative tuning for optimal ride feel and safety response
- Authorization flow: End-to-end testing from fingerprint scan to motor enable
- Reliability testing: Extended operation to verify no component failures under sustained use
Results
Successfully pivoted from Raspberry Pi to ESP32 without feature loss during global chip shortage.
Key Findings Project achievements
- Delivered stable, MVP-ready electronics system with full authorization flow
- Achieved target vehicle range of 50 km through motor controller optimization
- Successfully established CAN bus communication with VESC for comprehensive telemetry
- Pivoted from Raspberry Pi to ESP32 architecture during global chip shortage without feature loss
- Identified and resolved back-current issue using solid-state relay protection
- Created modular wire harness design enabling rapid prototyping and component swaps
Data Electronics iteration evolution table
Comparison across three development iterations
| Feature | Iteration 1 | Iteration 2 | Iteration 3 (MVP) |
|---|---|---|---|
| Microcontroller | Raspberry Pi Zero | RPi Zero (failed) | ESP32 |
| Programming Language | Python | Python | C/C++ |
| Control Interface | HW buttons + BT | Hardware buttons | Blynk IoT App |
| Harness Quality | Solid-core, loose | Improved lengths | JST, bundled |
| Motor Locking | Relay (unreliable) | Relay | Solid-state relay |
| Stand Detection | Hall effect | Hall effect | Hall (isolated) |
Iteration 3 achieved MVP status with all features functional and reliable for investor demonstrations.
Skills Technical skills demonstrated
- Embedded Systems: ESP32, Raspberry Pi, C/C++, Python
- IoT: Blynk IoT Platform, WiFi/BLE connectivity
- Communication: CAN Bus, UART, SPI protocols
- Hardware: PCB Design, hand soldering, SMD soldering, wire harness design
- Motor Control: VESC tuning, ESC configuration, telemetry parsing
- Sensors: Hall effect, fingerprint module, RPM sensor with ADC
Future Future work
- Implement RTOS for improved system responsiveness and real-time performance
- Convert battery voltage display to percentage-based indication in app
- Integrate RPM sensor as safety interlock
- Design professional PCB for production (replacing protoboard)
- Add GPS tracking and fleet management capabilities for shared mobility deployment
Team Contributions & credits
I designed the complete electronics architecture across 3 iterations including PCB design, ESP32 firmware, Blynk IoT app, and CAN bus motor controller integration. Led the platform pivot from Raspberry Pi to ESP32 during chip shortage.
Guided by Kithir Kareeb Navaz Ahmed Zackariya (Director & Co-Founder) and Harit Trivedi (Electronics Head) at DeltaX Automotive, with capstone guidance from Prof. Sachhidananda M H at PES University.
Mechanical chassis design and sensor mounting coordination done by the DeltaX mechanical team. Testing facilities provided by DeltaX Automotive.
Additional Information
Comprehensive internship report and presentation slides.
Presentation slides with visual summaries of key work areas.