As a CNRS spin-off, Amiral Technologies adapts to the automotive sector its algorithm for predictive maintenance and automatic features generation for test benches and vehicles.



Anticipate failures with factory 4.0

In the test benches and testing field, the prototypes tested are exposed to various hazards (failure, breakage, overheating, etc.). In fact, when a failure occurs, operators lose time to replace or repair the defective part. Knowing in advance the occurrence of a failure and its causes allows for greater monitoring, time and energy savings, as well as more accurate data.


An algorithm to reduce the downtime of test benches


Amiral Technologies has build with Route 26 a technical and commercial partnership to design a predictive monitoring solution for automotive test benches. The goal is to reduce the downtime of engine test benches during endurance tests.

This solution is based on an automatic features generation algorithm. Thanks to artificial intelligence, the algorithm learns through iterations to finally know in advance the failed component and the reason for its failure. As part of this collaboration, Route 26 tested and validated Amiral Technologies’ solution on its industrial engine test benches. This allowed the solution to be adapted to the automotive sector.

Thanks to the algorithm, the downtime of the benches can be reduced. This speeds up the process of validating innovations and the efficiency of test facilities by limiting intervention times.


Amiral Technologies performed a first PoC with EMC data. It was aimed to detect a drift indicating a failure on the catalyst thanks to data embedded in the engine. This PoC was performed with supervised learning on a failure set for an engine under the conditions of the given test (cycle type, engine type).


Amiral Technologies is specialized in Artificial Intelligence applied to Industrial Predictive Maintenance (defect prediction, aging, end-of-life estimation). It leverages a CNRS (French National Research Center) innovation that automatically extracts the characteristics determining the state of health of an equipment from the generated physical data. This allows for accelerated design of predictive models and unparalleled accuracy and performance. Amiral Technologies is a winner of the Digital Industry Program 2018 (organized by General Electric and NUMA) and a finalist in the AI Paris Region and Digital Industry Award challenges (organized by ATOS and Siemens). The startup is currently incubated at SATT Linksium Grenoble Alpes and accelerated in the Founders program at Station F in Paris.

See also

Retrofit of combustion engine