Big Data Processing Software Evolution for a Global Fleet Telematics Provider

About Our Client

The Client is a global fleet telematics provider that has been in business for over 25 years. The company's portfolio includes end-to-end telematics software products for fleet management and usage-based vehicle insurance. The Client continuously evolves its products to provide comprehensive fleet insights and convenient operations management experiences for its corporate clients.

Team Ramp-Up Needed to Drive Telematics Data Compatibility

At the core of the Client's products is a proprietary big data processing engine that enables real-time capture and analysis of fleet telematics data generated by vehicle devices. A back-end processing system then sends this data to the client-facing apps, enabling corporate users to monitor vehicle state, location, shipping conditions, and driver behavior.

As the Client's client base scaled, the company had to deal with a larger number of versatile fleet telematics devices. Disparate devices often relied on different telemetry protocols, which complicated the unified processing and centralized analysis of fleet event data. Plus, each client had specific fleet analytics needs and expected to obtain tailored analytics insights and alerts.

Accommodating each new telemetry protocol required the development of a dedicated data unification algorithm. As a result, the Client's in-house telematics system maintenance team had to handle an increasing load of continuous coding tasks. To alleviate the burden on its engineers and get the necessary solutions delivered quickly, the company decided to ramp up its team with outsourced talent. ScienceSoft came across as a trustworthy team augmentation vendor with solid experience in fleet management software development and specialized background in IoT and big data.

Development of Custom Telematics Data Unification Solutions

ScienceSoft provided a team of seasoned .NET developers who joined the Client's maintenance team. Our experts studied the available telematics software documentation, explored the vehicle devices the system commonly interacted with, and examined the supported raw and processed data formats. It helped them quickly grasp the specifics of the telematics data processing flows and understand potential compatibility issues.

In two weeks, ScienceSoft's engineers fully integrated into the Client's development processes and adapted to the team's communication model. Our experts' primary responsibility was building custom data standardization algorithms that would enable the following:

• Real-time capture of multi-format telematics data from vehicle devices (GPS trackers, accelerometers, fuel sensors, ELDs, truck temperature sensors, etc.).
• Automated data conversion to the Client's proprietary unified format designed for optimized telematics storage space and resource-friendly data transmission.
• Routing the unified, analytics-ready data to a database for storage. This data can be further used for event root cause analysis and forecasting.
• Instantly communicating data on the events of client interest (e.g., accidents, route changes, speeding over the limit, violations of the conditions of carriage, anomalies in driver behavior) to the relevant end users.

The Client's hardware team involved ScienceSoft every time they needed to onboard a new type of telematics device. Our experts investigated the device's data recording and transmission protocol, studied client requirements for the scope of reported events, and estimated timelines and resources to deliver a dedicated data unification solution.

ScienceSoft's team handled data unification protocol development end to end. They designed data capture and unification rules, coded the logic of the data processing algorithms, and integrated the algorithms into the telematics system's back end. Our engineers wrote automation test scripts and conducted functional tests using sample devices provided by the Client's customers. They ran QA procedures in parallel with coding to quickly spot any potential issues and prevent bugs in the live data processing system.

When end-user needs changed (e.g., a client wanted to obtain deeper telematics insights from its existing fleet device), ScienceSoft's developers evolved the corresponding data processing algorithm accordingly. They maintained detailed documentation on the delivered software to ensure prompt and smooth upgrades in the future.

Apart from development tasks, ScienceSoft's team helped the Client's engineers migrate the company's software product infrastructure (including the telematics processing system, client-facing apps, and databases) from Azure to Oracle Cloud. Also, the Client has recently reconsidered its approach to telematics data storage and involved ScienceSoft in assessing the feasibility of shifting from Microsoft SQL Server to Elasticsearch.

20 Custom Solutions to Standardize Data on 100B+ Fleet Events

As of May 2024, ScienceSoft has been cooperating with the Client for nearly three years. During this time, the Client received 20 robust data unification algorithms tailored to the specific telemetry protocols of various telematics devices.

Thanks to the mature Agile and QA practices and relevant domain expertise of ScienceSoft's talents, the Client got the opportunity to quickly adopt new telematics device types and meet its clients' specific fleet analytics needs. Since the start of our cooperation, the Client's telematics system has reached the capacity to recognize and process data on 100B+ vehicle events from hundreds of various devices.

Technologies and Tools

.NET, Microsoft Visual Studio, Microsoft SQL Server, RabbitMQ, Couchbase, Elasticsearch, Microsoft Azure, Oracle Cloud Infrastructure.