Scalable, AI-Powered TMS to Drive Market Expansion for a Logistics Company

Magma’s Transportation Management System (TMS) is a custom-built logistics platform designed to streamline operations across Asia and the US.

The solution includes multiple modules to ensure consistent management across managers, drivers, warehouse workers, etc. These modules enable intelligent routing, real-time tracking, barcode generation, and a mobile app, all integrated into a single comprehensive system. Built with efficiency and growth in mind, the TMS empowers Magma to outperform in the highly demanding competitive market.

Starting from producing labels for consignments to send parcels within the UK as a domestic market, Magma Trading evolved into a European-level logistics provider. When contacting us for collaboration, Magma attempted to enter the Asian and US markets alongside its partner companies, of course, with the help of Devox Software.

Magma had a demand for fast-paced development as the expanding process had already started. The project’s CEO picked Devox Software for our fast response and prompt decision-making, which was a must in terms of Magma’s request.

The project team included 16 members, including a Project Manager to oversee delivery, a Solution Architect for system design, and backend and frontend developers. Additionally, we invited AI/ML engineers to handle routing algorithms and part-time QA engineers for testing.

A DevOps engineer built cloud infrastructure, while a Business Analyst ensured that business requirements were implemented in the logic. UI/UX designers ensure usability and a sleek interface. This core team covered all key roles needed.

In 2019, during the development of the solution, we faced challenges due to the lack of robust AI-powered, ready-to-use solutions and libraries, which affected the implementation of advanced functionalities. Here are some of the challenges we faced crafting this TMS:

• Data Integration and Real-Time Tracking: Connecting numerous carriers, managers, drivers, and warehouse workers to create accurate and seamless data flow across different regions and networks.
• AI Algorithm Complexity: Developing and tuning routing algorithms to handle complex variables like traffic, weather, fuel consumption efficiency, and other delivery constraints.
• Scalability: The system must handle rapid growth across multiple countries with varying regulations and infrastructure.
• User Adoption: Training and onboarding drivers and staff on new digital tools and workflows.
• Latency and Connectivity: Overcoming connectivity issues in remote or less developed areas.
• Tight Timelines: We took part in expanding Magma’s solution, parallel to the product entering new markets.

Node.js for high-performance backend processing

React.js for a dynamic and responsive user interface

PostgreSQL for reliable, scalable database management

AWS Cloud Infrastructure to ensure availability and scalability

AI-driven routing algorithms to optimize delivery paths in real time

Python libraries like NetworkX

Zebra Crossing (ZXing) for Java or BarcodeLib for .NET

The high-paced client’s expectations required incremental development that would allow our joint team to deploy features as soon as they were designed and tested. Let’s focus on the modules we’ve been developing.

Delivery Management Module

The Delivery Management module streamlined logistics and optimized workflows:

• Real-time scheduling and rescheduling capabilities,
• Dynamic route adjustments based on traffic and weather conditions,
• Integration with partner delivery networks,
• Integration with third-party APIs, such as Google Maps, for route optimization and real-time tracking.

With a microservices architecture to ensure scalability and easy updates, Node.js for the backend, React for a responsive front end, and MongoDB for handling complex data structures, we met the bar of customer expectations.

Parcels Tracking Module

The Parcels Tracking module provides customers and partners with real-time visibility into parcel locations.

• Tracking functionality integrates IoT sensor data, mobile app updates from delivery personnel, and GPS data, creating a unified system that provides location updates and status changes for each parcel,
• Apache Kafka enabled real-time data streaming and Elasticsearch for fast, complex data querying, providing a robust infrastructure capable of handling high traffic as Magma entered new markets,
• React Native for a cross-platform mobile application interface for end customers and staff to access tracking updates.

Furthermore, the module evolved with iterative feature expansions, such as estimated delivery time predictions and customer notifications, based on Magma’s ongoing needs.

Returns Management Module

To complete the logistics cycle, Devox developed the Returns Management module, creating a streamlined system for handling customer returns and inventory updates.

• User-friendly interfaces for initiating returns, tracking return status, and processing refunds or exchanges,
• The backend used Java with Spring Boot or Node.js for managing return workflows, integrated with MySQL to handle the large volumes of return data securely,
• Automated processing rules, built with BPMN standards, helped streamline decision-making for refund eligibility and product restocking,
• Automated workflows triggered notifications to customers and updated the inventory upon return, reducing manual work for Magma’s logistics team.

Intelligent Routing Module

The Intelligent Routing module transformed Magma’s logistics cost and time efficiency by optimizing delivery routes across different geographies:

• Algorithms for the shortest path helped implement an AI-powered routing engine with Python libraries like NetworkX for network analysis and TensorFlow for predictive analytics,
• Real-time traffic, weather, and delivery demand data fed the system, dynamically adjusting routes and reducing fuel consumption and delivery times,
• Sophisticated route optimization was enabled by Google OR-Tools and custom genetic algorithms.

Barcode Generation Feature

With our help, the system started dynamically generating and managing barcodes for efficient tracking and processing:

• Libraries like Zebra Crossing (ZXing) for Java or BarcodeLib for .NET allowed the client to create unique barcodes associated with each parcel,
• Managed with Spring Boot (Java) or Express.js (Node.js) for robust data handling, barcodes needed a reliable PostgreSQL database,
• Integrated error-correction algorithms ensured barcode readability and were tested in QA sprints for high accuracy under real-world conditions.

Andy Morrey, a Managing Director at Magma Trading, stated:

“I’m blown away by the level of professionalism that’s been shown, as well as the welcoming nature and the social aspects. You all seem to be really on the ball technically.”

With the help of Devox, Magma Trading managed to:

• Expand to two significant new markets simultaneously (Asia and the USA) while scaling the product,
• Roll out new features swiftly and on the go with their market expansion,
• Achieve +45%* to operational efficiency by automating key processes, reducing manual intervention, and optimizing delivery routes,
• Ensure a scalable and resilient system infrastructure that could accommodate future growth without compromising performance.

*According to the internal audit