SafeRoute AI: Smart logistics solution for real-time truck tracking
Challenge
Our client, a major logistics and transport company with an international fleet of trucks, faced a growing challenge: gaining actionable insights from large volumes of raw telematics data to improve safety, optimize routes, and reduce vehicle wear and tear. While the trucks were equipped with Teltonika FMB920 devices transmitting GPS, accelerometer, and CAN bus data, the client lacked a centralized system to analyze this data efficiently.
The previous process relied heavily on manual data aggregation or expensive streaming platforms, which either slowed down decision-making or led to overcomplex infrastructure costs. Moreover, there was no real-time route deviation detection, behavior monitoring, or integration with existing ERP, TMS, or WMS systems. The client needed a cost-effective, scalable, and DBMS-free solution that could provide structured insights from high-frequency telematics data while working in environments with limited computing resources.
Our client, a major logistics and transport company with an international fleet of trucks, faced a growing challenge: gaining actionable insights from large volumes of raw telematics data to improve safety, optimize routes, and reduce vehicle wear and tear. While the trucks were equipped with Teltonika FMB920 devices transmitting GPS, accelerometer, and CAN bus data, the client lacked a centralized system to analyze this data efficiently.
The previous process relied heavily on manual data aggregation or expensive streaming platforms, which either slowed down decision-making or led to overcomplex infrastructure costs. Moreover, there was no real-time route deviation detection, behavior monitoring, or integration with existing ERP, TMS, or WMS systems. The client needed a cost-effective, scalable, and DBMS-free solution that could provide structured insights from high-frequency telematics data while working in environments with limited computing resources.
Solution
WEZOM developed a batch-processing fleet analytics platform that aggregates GPS, accelerometer, and CAN bus data from Teltonika FMB920 trackers into a unified JSON format. These messages are processed in Python and converted to Excel flat files, allowing non-technical users to access reports without a dedicated database or complex streaming architecture.
We built algorithms for detecting idle time, route deviations from GPX tracks, speeding (based on both GPS and CAN metrics), and aggressive driving (via accelerometer). Additionally, the system supports ADAS event scenarios like lane departures or collision warnings. All relevant events are logged and visualized via Excel or can be accessed through API for external system integrations.
To ensure seamless data exchange, we implemented REST API and MQTT broker support, enabling integrations with ERP, TMS, WMS systems, and external logistics platforms. The architecture was designed to work offline or in constrained environments and prioritized simplicity, extensibility, and low infrastructure costs—perfectly meeting the client’s operational constraints and business goals.
WEZOM developed a batch-processing fleet analytics platform that aggregates GPS, accelerometer, and CAN bus data from Teltonika FMB920 trackers into a unified JSON format. These messages are processed in Python and converted to Excel flat files, allowing non-technical users to access reports without a dedicated database or complex streaming architecture.
We built algorithms for detecting idle time, route deviations from GPX tracks, speeding (based on both GPS and CAN metrics), and aggressive driving (via accelerometer). Additionally, the system supports ADAS event scenarios like lane departures or collision warnings. All relevant events are logged and visualized via Excel or can be accessed through API for external system integrations.
To ensure seamless data exchange, we implemented REST API and MQTT broker support, enabling integrations with ERP, TMS, WMS systems, and external logistics platforms. The architecture was designed to work offline or in constrained environments and prioritized simplicity, extensibility, and low infrastructure costs—perfectly meeting the client’s operational constraints and business goals.
Results
The SafeRoute AI platform successfully transformed raw truck telematics into actionable insights. The client now receives structured reports on route deviations, speeding, idle time, and aggressive driving behavior—without needing real-time streaming infrastructure or a full-fledged database.
By processing everything in batch mode using Excel-compatible outputs, the solution enabled wider adoption across departments and empowered non-technical stakeholders with clear, digestible insights. The high-frequency tracking data (up to 10 Hz) is now reliably aggregated and analyzed daily, ensuring accurate logistics optimization.
The integration of REST APIs and MQTT brokers also allowed the client to connect their ERP and TMS platforms to the system in real time, ensuring synchronized workflows and faster decision-making. Moreover, support for ADAS event scenarios helps the company proactively manage safety risks. This low-cost, scalable, and highly effective fleet tracking analytics system is now a core part of the client’s digital logistics strategy.
The SafeRoute AI platform successfully transformed raw truck telematics into actionable insights. The client now receives structured reports on route deviations, speeding, idle time, and aggressive driving behavior—without needing real-time streaming infrastructure or a full-fledged database.
By processing everything in batch mode using Excel-compatible outputs, the solution enabled wider adoption across departments and empowered non-technical stakeholders with clear, digestible insights. The high-frequency tracking data (up to 10 Hz) is now reliably aggregated and analyzed daily, ensuring accurate logistics optimization.
The integration of REST APIs and MQTT brokers also allowed the client to connect their ERP and TMS platforms to the system in real time, ensuring synchronized workflows and faster decision-making. Moreover, support for ADAS event scenarios helps the company proactively manage safety risks. This low-cost, scalable, and highly effective fleet tracking analytics system is now a core part of the client’s digital logistics strategy.
