Drivora | Distributed Universal Advanced Driver Assistance System

01 / Introduction

The Problem We're Solving

Many legacy vehicles lack modern ADAS - leaving drivers without collision warnings, blind-spot alerts, lane assistance, or stability monitoring.

Drivora is a distributed retrofit safety system that brings comprehensive driver assistance to any vehicle at a fraction of the cost of a new car.

Blind Spots

Large vehicles have massive blind spot zones that cause accidents during low-speed maneuvering.

Lean & Stability Risk

Unsafe leaning or sudden vehicle movement can create risky driving conditions, especially during turns and quick maneuvers.

Legacy Vehicles

Many older vehicles still do not include collision warning, blind-spot monitoring, lane assistance, or real-time safety visualization.

Target Audiences

Older Passenger Vehicles

Owners of classic and older vehicles who want modern safety features without the cost of purchasing a new vehicle.

No collision warning system
Missing blind spot detection
Limited safety budget

Commercial Vehicles

Operators of buses, lorries, and delivery trucks with massive blind spots and high centers of gravity.

Massive blind spot zones
High unsafe lean conditions
Low-speed collision risk

Key Safety Features

Forward Collision Warning

Front ultrasonic sensing and camera-based object detection work together to identify possible frontal obstacles.

Blind Spot & Rear Safety

Three rear ultrasonic sensors monitor left, center, and right rear zones for blind-spot and reverse safety alerts.

Lane Departure Warning

Raspberry Pi camera-based road marking detection with immediate audible alerts to support lane keeping.

Lean & Stability Monitoring

IMU-based lean and tilt monitoring to warn the driver about unsafe vehicle movement conditions.

02 / Solution Architecture

System Architecture

Drivora is built as a distributed system with four hardware modules communicating over a CAN-Bus backbone, all connected to a Raspberry Pi brain unit that handles vision processing, sensor fusion, decision logic, dashboard hosting, and audio alerts.

Prototype Implementation

Component-level view of the physical prototype, showing how each hardware module is wired and integrated within the vehicle.

Data & Control Flow

Sensors

Front/rear ultrasonic sensors, IMU, and Raspberry Pi camera collect environmental and vehicle motion data.

→

CAN-Bus

Sensor units transmit data through the CAN-Bus backbone to the Raspberry Pi brain via an ESP32-C3 CAN bridge.

→

Processing

The Raspberry Pi fuses CAN sensor data with camera processing results and applies decision logic to generate warnings.

→

Output

Audio alerts and real-time visual warnings are shown through the local smartphone dashboard.

Communication Methods

Protocol	Used For	Direction
CAN-Bus	Inter-module communication backbone	Bi-directional
Local Wi-Fi	Smartphone app user interface	Hub → Phone
Internet Sync	Cloud event logging and analytics	Brain → Cloud

03 / Hardware & Software Designs

Detailed System Designs

Hardware Design

Sensors

Sensor	Model	Qty	Application
Front Ultrasonic	JSN-SR04T 3.0 (Waterproof)	2×	Forward collision detection
Rear Ultrasonic	JSN-SR04T 3.0 (Waterproof)	3×	Rear / blind-spot coverage
IMU	MPU-6050	1×	Lean & stability monitoring
Camera	Raspberry Pi Camera Module 3	1×	Lane detection & object detection

Processing Units

MCU

Raspberry Pi 3 Model B - Brain Unit

Primary processing unit handling camera processing, sensor fusion, decision logic, local Wi-Fi hotspot dashboard hosting, and audio warning control. This is the brain of the Drivora system.

MCU

ESP32-C3 - Sensor Unit Controllers and CAN Bridge

ESP32-C3 boards are used in the front unit, rear unit, center lean unit, and CAN bridge to handle local sensor reading and CAN-Bus communication.

Communication Hardware

CAN Transceivers

4× TJA1050 CAN modules enable reliable communication between the distributed sensor units and the Raspberry Pi brain through the CAN bridge.

Power Systems

Buck converter and adapter ensure stable power delivery to all system components from the vehicle's 12V power rail.

Enclosures & Wiring

3D printed housings and waterproof connectors protect the outdoor front and rear units while supporting secure vehicle installation.

Software Design

Raspberry Pi Brain Software

CAN Data Reception
via ESP32-C3 bridge

→

Sensor Fusion
Data normalization

→

Decision Logic
Threshold evaluation

→

Alert Generation
Audio / Web Dashboard

→

Local Dashboard
Smartphone view

Smartphone Dashboard

Live Safety Dashboard

Real-time display of all sensor readings and safety status indicators.

Alert Notifications

Immediate visual and audio warning indications for critical safety events.

Critical Event Logs

Recent warning and sensor states shown in a simple driver-focused interface.

System Configuration

View unit status and basic system state through the local dashboard interface.

Future Data Logging

The current prototype focuses on real-time local warnings, while future development can add long-term logging and analytics. Possible events to log include:

High Lean Events Lane Departures Forward Collision Warnings Blind Spot Triggers Lean & Stability Risk Events System Faults

Enclosures

Custom 3D-printed waterproof housings designed to protect each sensor unit in outdoor vehicle-mounted conditions.

CAD Designs

Front Unit Housing

Rear Unit Housing

Brain Unit Enclosure

Lean Unit Housing

Physical Prototypes

Front Unit

Rear Unit

Brain Unit

Lean Unit

04 / Testing

Testing & Validation

Drivora follows a multi-level testing strategy, from isolated sensor validation through integration, vehicle fitment, and real-world prototype testing.

Module Testing

Validate each hardware module independently using simulated inputs before system integration.

⬤

Front Ultrasonic Distance Accuracy

Verify ultrasonic measurements against known reference distances at various ranges.

⬤

Ultrasonic Echo Timing

Validate round-trip echo timing and distance computation accuracy.

⬤

IMU Tilt / Vibration Response

Test MPU-6050 readings under controlled tilt and vibration conditions.

⬤

Camera Lane Detection

Evaluate Raspberry Pi camera-based lane detection under varied lighting conditions.

Integration Testing

Test CAN-Bus communication and system-level functionality with all modules connected.

⬤

CAN Data Transmission

Verify data integrity and latency across the CAN-Bus network under load.

⬤

Sensor Integration Accuracy

Validate sensor fusion output against combined sensor scenarios.

⬤

Alert Response Time

Measure end-to-end latency from sensor trigger to driver alert generation.

⬤

Error Handling

Validate graceful degradation when a module fails or disconnects.

Real-World Testing

Validate the complete Drivora system in live traffic conditions on public roads.

⬤

Forward Collision Warning

Test warning generation at realistic approach speeds in controlled traffic scenarios.

⬤

Blind Spot Detection

Validate detection accuracy for vehicles in blind spot zones during lane changes.

⬤

Lane Departure Accuracy

Evaluate false positive/negative rates across road types and weather conditions.

⬤

Lean Warning Validation

Test lean warning behavior during controlled vehicle turns and movement conditions.

Project Timeline

Week 17

Design & Procurement Wk 1 – 4

Module Development Wk 4 – 10

Integration & Software Wk 11 – 16

Testing & Refinement Wk 17 – 20

Wk 1Wk 5Wk 10Wk 15Wk 20

05 / Gallery

Project Gallery

A visual record of our build process - hardware assembly, testing sessions, and prototype integration milestones.

05 / Budget

Detailed Budget

Rs. 30,502 Updated Prototype Budget

Updated project cost based on the latest component, wiring, connector, courier, and 3D printing expense list.

Grouped Expense Items

#	Budget Group	Included Items	Total (Rs.)
1	Sensor & Signal Modules	CDM324 radar sensor, MPU-6050 IMU, JSN-SR04T ultrasonic sensors, LM358 amplification modules	14,142.00
2	Controller & CAN Communication Modules	ESP32-C3 Super Mini boards, TJA1050 CAN modules, MCP2515 CAN module	3,770.00
3	Power & Protection Components	XL4016 buck converter, 1N5408 diodes, glass fuse	780.00
4	Wiring, Connectors & Cable Accessories	Jumper wires, twine wire, waterproof connectors, heat shrink tubes, cable ties, 4-core wires	4,440.00
5	Mechanical Mounting & 3D Printed Parts	Brass threaded inserts, M3 screws and nuts, front unit 3D print, rear unit and main hub mounts	6,290.00
6	Courier Fees	Courier charges for component orders	1,080.00
Grouped Expense Total			30,502.00

Final Cost Summary

Detailed Expense Items Total	Rs. 30,502.00
Additional Adjustment	Rs. 0.00
Final Prototype Budget	Rs. 30,502.00

* Similar items have been grouped to keep the budget section concise while preserving the total from the latest expense table.

06 / Conclusion

Conclusion & Future Work

What Was Achieved

Drivora demonstrates a viable, low-cost approach to retrofitting modern ADAS capabilities onto any vehicle. By distributing sensor units across the vehicle and connecting them via a CAN-Bus backbone, the system provides front collision warning, rear blind-spot and reverse safety monitoring, lane departure alerts, lean monitoring, and real-time dashboard visualization.

The updated prototype budget of Rs. 30,502 includes the main electronic components, sensor modules, wiring, waterproof connectors, courier charges, and 3D printed enclosures used during prototype development.

Future Developments

AI-based Threat Assessment
Improve camera-based object detection and fusion logic for smarter collision prediction and false positive reduction.
Fleet Management Cloud Platform
Develop a dedicated cloud dashboard for fleet operators to monitor multiple vehicles in real time.
OTA Firmware Updates
Improve remote update and maintenance support for the Raspberry Pi brain and ESP32-C3 sensor units.
V2V Communication
Explore additional vehicle-to-vehicle or cloud-assisted safety features for future versions.

Commercialization Plans

Pilot Program

Partner with vehicle owners or small fleet operators for real-world prototype trials and data collection.

Certification

Improve hardware reliability, enclosure durability, and installation safety before larger-scale deployment.

Product Launch

Refine Drivora into an aftermarket safety add-on that can be installed through vehicle service centers.

Regional Scale

Scale the solution for wider use in regions where many vehicles still lack modern safety systems.

"Every vehicle deserves a second chance at safety."

The Problem We're Solving

Blind Spots

Lean & Stability Risk

Legacy Vehicles

Target Audiences

Older Passenger Vehicles

Commercial Vehicles

Key Safety Features

Forward Collision Warning

Blind Spot & Rear Safety

Lane Departure Warning

Lean & Stability Monitoring

System Architecture

Prototype Implementation

Data & Control Flow

Sensors

CAN-Bus

Processing

Output

Communication Methods

Detailed System Designs

Hardware Design

Sensors

Processing Units

Raspberry Pi 3 Model B - Brain Unit

ESP32-C3 - Sensor Unit Controllers and CAN Bridge

Communication Hardware

CAN Transceivers

Power Systems

Enclosures & Wiring

Software Design

Raspberry Pi Brain Software

Smartphone Dashboard

Future Data Logging

Enclosures

Drivora Flutter App

Testing & Validation

Module Testing

Integration Testing

Real-World Testing

Project Timeline

Project Gallery

Detailed Budget

Grouped Expense Items

Final Cost Summary

Conclusion & Future Work

What Was Achieved

Future Developments

Commercialization Plans

Pilot Program

Certification

Product Launch

Regional Scale

Meet the Team

Project Members

Thakshila Bandara

Prasadi Bandara

Shamishka Darshana

Sachith Nirmal

Project Supervisors

Ms. Yasodha Vimukthi

Dr. Isuru Nawinne

Mr. Thilina Gunarathne