MRail IntelligenceMARIC · proof of concept
RailMarket.com · 02 November 2026 · en

„In railway telematics, AI is not always the solution“, says Eyal Trichter, CEO of Trilogical Technologies

Read the original ↗

Summary

AI · gpt-5.4-mini-2026-03-17 · relevance 45

Trilogical Technologies CEO Eyal Trichter said railway telematics is shifting from data collection to operational decision support. The company is investing in sensors, firmware, software and cloud analytics for freight wagons and passenger rolling stock, with APIs for integration into operators’ existing fleet systems.

Malaysia’s rail suppliers should note growing demand for interoperable telematics, onboard sensing and analytics platforms that can integrate with operators’ maintenance and fleet systems.
Ringkasan

Ketua Pegawai Eksekutif Trilogical Technologies, Eyal Trichter, berkata telematik rel kini beralih daripada pengumpulan data kepada sokongan keputusan operasi. Syarikat itu melabur dalam sensor, perisian tegar, perisian dan analitik awan untuk gerabak barang serta stok rolling penumpang, dengan API untuk integrasi ke sistem armada sedia ada operator.

Extracted text (15299 chars)
RM: When we spoke two years ago, a large part of the discussion was about collecting data from trains and freight wagons. At InnoTrans 2026, you are talking much more about turning this data into operational decisions. What has changed? Eyal Trichter, Trilogical: During the last two years, we focused on developing technologies that differentiate us in the areas where we operate. We provide a cross-platform solution that can work across different OEMs and fleet owners, and at the same time we collect very large amounts of raw data - billions of records. AI has changed what we can do with this information. Two years ago, producing useful outcomes from such volumes of data was much more challenging. Today, access to cloud-based information is easier and developing advanced analytical models has become faster. This combination of high-quality data and advanced analytics means that we can show customers results much earlier. For example, we can take a year of historical data containing billions of records, run a model against it and demonstrate what a certain condition or analytical rule would have detected during that period. RM: Has most of the development therefore moved towards software and analytics, or are you still investing significantly in onboard hardware? Eyal Trichter, Trilogical: Both. We have two main product areas. RailBlazer is our solution for smart freight wagons, while Control FREAK is used for locomotives, multiple units and passenger rolling stock. With RailBlazer, we invested considerably in the hardware. We developed new sensors and worked on calibration to make the measurements more precise and better suited to customer requirements. On the locomotive and multiple-unit side, much of the work has been in firmware, software and the backend. We are making the onboard devices smarter and developing models that can process and analyse data in real time. A large part of our investment has also gone into the cloud platform and the way this information is presented to operations, maintenance and engineering teams. RM: What does that mean in practice for these three groups? Eyal Trichter, Trilogical: Operations can see an issue during a live trip and immediately access the information needed to make a decision. They can decide what action should be taken, what information should be passed to the driver or whether the train needs to stop. For maintenance, we provide a health indication for individual assets. The system can identify which equipment requires higher maintenance priority. The purpose is to support reliability and availability by directing attention to the assets where it is needed. Engineering has a different requirement. Engineers want to investigate a failure, compare it with similar events and understand the conditions in which it occurred. They can develop a new analytical condition and test it on historical data before applying it to future operation. This allows them to see how such a condition would have behaved in the past and whether it could provide an earlier warning of a critical failure. RM: Railway companies already have their own fleet management and maintenance systems. Do they normally use your platform directly, or do you feed Trilogical data into systems they already operate? Eyal Trichter, Trilogical: We can do both. Many customers want to continue using an internal system they already know, so we provide APIs and they can integrate our information into their own platform. At the same time, our own platform provides much deeper analytical capabilities. For operators, maintainers, engineers and managers who want to investigate the data in detail, using our platform gives them access to more functions. We divide it according to the requirements of operations, maintenance and engineering, but these groups can also share information and insights. That is important because the platform gradually becomes a knowledge base. If one engineering department establishes, for example, the parameters…

Entities

Companies
Topics
rail telematicsAI analyticspredictive maintenancefreight wagonscloud platforminteroperability