Smartphone-first AI safety for West African roads

Empowering Drivers. Securing Fleets. Saving Lives.

The first AI-powered, hardware-free telematics and highway security platform built specifically for West African roads.

Download the Driver App Workshop Questionnaire Access Partner Dashboard

A globally recognized research initiative backed by leading academic institutions and driving institutional change alongside the FRSC and NURTW.

The Challenge

Traditional vehicle trackers are expensive, require complex hardware installations, and act as punitive surveillance tools rather than proactive safety coaches. Most telematics systems are designed for Western, high-income settings and fail to account for Nigerian realities like inconsistent internet, sparse speed-limit data, and local driving culture. Furthermore, commercial drivers on high-risk interstate routes face severe security threats like highway robbery and hijacking without an automated lifeline.

The Safe Drive Africa Solution

We transform the smartphone already in the driver's pocket into an advanced AI telematics device. We provide an offline-capable system that detects unsafe habits, delivers culturally attuned and legally grounded coaching, detects likely alcohol-influenced driving behaviour on-device, provides a partner analytics hub, and features automated hijack detection to keep drivers secure. Zero hardware required.

Core Platform Features

Safety intelligence, impairment detection, and security without expensive hardware.

Hardware-Free Telematics

No costly hardware installation. Our app captures GPS and inertial smartphone sensor data to detect harsh braking, severe swerving, and speeding, functioning seamlessly even in offline or low-connectivity environments.

Culturally Attuned AI Coaching

Generic safety alerts are ignored. Our dual-component Natural Language Generation (NLG) engine delivers localized safety tips and reflective weekly reports that speak to the driver's culture. Tips include legally grounded reflections explicitly citing Nigerian traffic laws and specific penalties.

Highway Security & Hijack Alerts

Built for the realities of the road. Utilizing intelligent existing sensor data to model high G-Jolt events, vanished sensor vibrations, and abrupt vehicle motion state changes, the app instantly alerts fleet managers and transport unions to suspected hijackings or severe accidents.

On-Device Alcohol Impairment Detection

Our app features an integrated, on-device decision-tree machine learning classifier that detects likely alcohol-impaired driving without the need for breathalyzers. This system analyzes real-world sensor-derived driving temporal and variability patterns to establish high-precision classification.

Technology & Impact

Proven Technology. Measurable Impact.

The partner dashboard mirrors a live control room: fleet safety operations, road-safety oversight, insurance risk analytics, security escalation, and driver improvement data in one operational view.

Fleet Control Simulation

Lagos to Abuja Corridor

Live

Active vehicles

83%

Safe driving

Priority alerts

ABJ-14

KAD-07

LAG-22

Privacy by Design

We track safety, not every move. The app uses Activity Recognition to automatically detect trips, processing data locally on the device to conserve battery and protect driver privacy.

Proven Behavioural Improvement

In a within-subjects evaluation pilot, Unsafe Behaviours Per Kilometre (UBPK) decreased by 19.4%, with 79.6% of participating drivers showing measurable improvement.

90.91% Accurate Alcohol Detection

An on-device decision-tree machine learning classifier detects likely alcohol-impaired driving without the need for breathalyzers.

Data-Driven Precision

Trained on a Nigeria-collected dataset, the model reaches 100% recall, 75% F1 score, 90.91% accuracy, and an AUC of 0.855.

Real-Time Analytics

Access a comprehensive, bird's-eye view of your entire fleet's safety metrics without paying for GPS hardware trackers.

Risk Scoring & Insurance

Use concrete data on driver improvement and reduced collision risks to negotiate better premiums with insurance providers.

Instant Security Routing

Receive immediate, high-priority notifications if a device detects motion anomalies consistent with a hijack or severe accident.

Launch Partner Dashboard

Co-Development

From Laboratory to the Highway: Creating Real-World Impact

We do not build technology for stakeholders; we build it with them.

FRSC

Federal Road Safety Corps

Aligning AI coaching algorithms with the Nigeria Highway Code to support national crash reduction mandates.

NURTW

National Union of Road Transport Workers

Partnering with driver unions to ensure Hijack Detection features protect members on high-risk routes.

Cross-Border Logistics

Regional transport continuity

Collaborating with Cameroonian transport officials and international haulage operators to support cross-border safety.

Institutional Research

The Research Team Behind the AI

TETFund and NDU origin

Safe Drive Africa's core technology, mobile app, and AI models were originally developed as a doctoral research program fully funded by the Tertiary Education Trust Fund (TETFund) through Niger Delta University (NDU).

IEAF impact acceleration

Today, its transition into a scalable, real-world institutional tool including the B2B dashboard, hijack detection, and stakeholder engagement is backed by an Impact Case Study grant from the University of Aberdeen's Impact and Engagement Accelerator Fund (IEAF).

Dr. Iniakpokeikiye Peter Thompson

Primary Researcher & Innovator

The architect of the Safe Drive Africa platform. Dr. Thompson designs and evaluates smartphone-first AI systems that turn real-world mobility data into actionable, culturally resonant safety feedback in low-resource environments.

Visit Personal Website

Dr. Dewei Yi

Project Lead

An Associate Professor of Robotics and AI at the University of Aberdeen, Dr. Yi spearheads the institutional impact strategy. His expertise in machine learning and autonomous vehicular networks helps translate the platform into data-driven outcomes that lower operational costs.

View Research Profile

Professor Ehud Reiter

NLG Expert

A global pioneer in Natural Language Generation at the University of Aberdeen. Prof. Reiter's foundational academic backing drives the platform's unique ability to translate raw sensor data into persuasive, localized safety coaching that changes human behavior.

Read Prof. Reiter's Blog

Peer-Reviewed Scientific Foundations

An End-to-End System for Culturally-Attuned Driving Feedback using a Dual-Component NLG Engine (Thompson, I. P., Yi, D. & Reiter, E., IEEE)

Mobile Phone Sensor-based Nigerian Driving Dataset to Detect Alcohol-influenced Behaviours (ICAC)

Safe Drive Africa: Culturally Attuned AI-Enabled Road-Safety Intervention for Nigeria

Download the Safe Drive Africa driver app.

Install the Android APK on a phone or compatible device. Fleet operators, government road safety officials, insurance partners, and institutional collaborators can use the same app package for approved pilots.

Download APK