Turning innovative ideas into success stories.
We use sensor data analytics to avoid costly equipment failure and improve operational efficiency.
Getting business insights from historical performance data has became a wonderful vector of
growth for many companies.
New technologies such as machine learning, artificial intelligence and big data show promising
results in the field of predictive maintenance.
However in practise, they often fail to capture nuances that are obvious for human.
Liaising with the technical teams in the field, our approach is to combine advanced data science techniques together with domain experts knowledge in order to reliably extract those characteristic patterns. Empowering people with our condition monitoring algorithms enables them to forecast equipment failures before they occur. This helps our partners save time and money in their daily operations.
We blend in an incredibly intuitive user interface, the most advanced optimization algorithms.
Within an organization, system-level model design and calibration requires a know-how that is
often in the hand of a very small team.
As the need for simulation spread within those companies,
they often end up spending more time maintaining their current library
rather than developing new ideas.
We designed a new development workflow where each model is associated with a dedicated semi-automated calibration routine. First, thanks to a versatile user interface, experts build an easy-to-share model. They also define a matching calibration process by combining elementary optimisation stages. The model is then shared as a single package that allows the end user to calibrate and run the model seamlessly. A dedicated simplified user interface ensures that this can be achieved without prior simulation or optimization knowledge, thus participating in promoting the use of simulation throughout the company.
We empowered engineers with a tailored software allowing them to become more agile.
The automotive industry faces a unique challenge whereby the requirement for technological
is dictated not only by a fierce competition but also by an ever more stringent set of
Working closely with a car manufacturer engineering department, we developed a tailored tool suite to predict the performance of their upcoming powertrains. By integrating the brand process and methods and the engineers know-how directly in our software, we enabled them to estimate the pollutant emissions level earlier than ever in their development workflow. Engineers can now easily narrow down the most critical operating conditions where to focus their efforts. This eventually reduces the time and cost required to converge toward a viable solution. This project became a partnership towards continuously minimizing the requirements for test bench trials while maintaining the same level of predictability.
We use digital twins to help operators take the right decision.
In many industries, data intelligence has became a major vector of growth in the past few years.
By working closely with engineers and technicians on site and sharing the know-how we acquired
in the field of engineering digitalization,
we help our partners aggregate large amount of historical and live data to generate valuable
Using the data of sensors already available, together with some of the advances we brought in other industries, we were able to combine the benefits of physics-based and statistical simulation models into advanced digital twins. Those are live tuned using the latest data available and used to provide live advice based on accurate predictions. In order to provide support to the commercial teams, we developed a fully functional offline version of the tool that demonstrates the potential benefits and helps get new custmers on board. Combining data fusion algorithms with a model predictive control approach, we help companies improve their operational efficiency.
We want to bridge the gap between simulation and embedded software.
Engineers in the automotive industry have to cope with the growing complexity of systems
and software due to the introduction of advanced functionalities towards minimizing real driving
(e.g. hybrid powertrains, electrification), optimal operation strategies and/or automated
In the EMPHYSIS project (EMbedded systems with PHYSical models In the production code Software), we embbed physics-based models into the ECU's control and diagnosis functions. With that approach and our partners, we aim to trigger genuine enhancements in automotive vehicle performances within the project life cycle. This is a critical milestone towards reducing the development cost and time of embedded systems software but also contributes to enhance the overall performance of the production code in automotive vehicles. Learn more
We are active on several research projects through the development of leading edge numerical algorithms.
It is important for us to invest time in technologies which strive for positive social impacts.
Since 2015 we work in partnership with the French laboratory CNRS-PROMES at the Themis solar
on concentrated solar irradiation. This technology is one of the most promising value-creation
opportunities in desert areas.
Thanks to advances in data and image analysis, we helped the researchers get a better understanding of the solar flux concentration phenomenon so that they can obtain a live and accurate image of the solar flux received. In particular this led to the development and publication of an innovative approach to quantify the distribution of concentrated solar irradiation. Using this new measurement technique allows to develop more efficient control strategies, a significant leap towards making it a viable renewable energy production aternative. Learn more