- RESEARCH & INNOVATION FOR MAINTENANCE -
CARL SOFTWARE | BERGER LEVRAULT
CARL SOFTWARE | BERGER LEVRAULT
CARL Software, an expert in equipment management solutions and / or EAM  for over 30 years, has the largest specialized CMMS  team in Europe and also relies on a worldwide network of distributors.
Business sector: Computer software
Type of building concerned: community, administration, tertiary buildings, industries, infrastructures.
“We want to develop a global, universal, intelligent and low-cost IoT solution (economically adapted to the scale of a building) that integrates Artificial Intelligence and Edge Computing, based on information processing that can be split between Cloud processing as well as embedded in the physical sensor”.
Located respectively at the two ends of the IoT data value chain, the Adeunis sensor reads the data on one hand and transmits it to the CARL Software’s IoT platform on the other hand, which analyses and integrates it into the maintenance processes. Data processing is optimized by Edge Computing at the edge of the network.
“Our credo is to simplify the life of our customers and offer new uses at the cutting edge of innovation with this predictive maintenance service that is simple to manage and deploy, and which frees itself from technical obstacles with a high economic impact thanks to Edge Computing”, explains Youssef Miloudi.
“Through the integration of maintenance algorithms and decision elements in Adeunis sensors and the use of Edge Computing, we will enable our customers to save computing time, storage space, security risks and ultimately facilitate their operation and responsiveness, and reduce the cost of implementing predictive maintenance solutions”, explains Youssef Miloudi.
The Adeunis / CARL Software integrated offer adds more intelligence into Adeunis sensors and facilitates the implementation of Artificial Intelligence algorithms in the CARL Software IoT platform to meet the following objective: prevent and anticipate the technical maintenance of equipment’s through the generation of predictive models whose role is to detect equipment malfunctions or drifts over time.
“We’re planning two floors to our solution. The first is the development of an “augmented” product that embeds the ability to handle multiple physical modalities in flexible, cloud-driven time windows. We are applying advanced mathematical processing to these windows.
The second stage consists of directly embedding AI and malfunction detection algorithms that have been learned in the Cloud”, explains Jean-Luc Baudouin.
The pilot project:
Initially, the two entities chose to rely on the Adeunis Delta P sensor. This sensor, who is in the CARL Software building, monitors the proper functioning of the ventilation systems. The intelligence provided will make it possible, notably through cycle analyses, to anticipate system maintenance and gain a better understanding of the failures detected. This first project serves as the basis for the creation of a platform of embedded algorithms common to all Adeunis sensors.
“This project is the first brick of a structuring program for Adeunis. The technological evolutions of silicon and the rapid development of Artificial Intelligence make it possible to carry out treatments that would have been unthinkable just a few years ago. This opportunity, coupled with the possibility of collaboration with the Cloud, offers exceptional prospects for IoT and the use our customers make of it.
We are on our way to build the technical foundation and new skills that will enable us to position ourselves as a leader in this revolution“, enthuses Jean-Luc Baudouin.
Technical and economic benefits for maintenance
Directly integrated into the CARL Source CMMS, Adeunis connected devices are able to create maintenance orders without any intermediary, based on onboard analyses! Unlike traditional IoT or supervision systems, it is no longer necessary to set alert thresholds and exchange interfaces; the “Adeunis – CARL Source connected object” cooperation is transparent and automatic!
Thanks to the relevant and more precise information transmitted, the maintainer can anticipate his needs or improve his responsiveness, better target his maintenance actions, reduce intervention costs and act on the energy performance, maintenance costs and durability of the equipment.
Anticipation of needs and increased reactivity for the maintainer
Better targeted maintenance actions
Reduced intervention costs
Technical incubation phase and development of the solution
Real tests on the CARL Software pilot building
Provision of a joint Adeunis / CARL Software offer
Data analysis for predictive maintenance requires the collection of a quantity of information at variable sampling frequencies depending on the dynamics of the system (on the order of one second for a rotating system driven by an electric motor per hour for a system with high thermal inertia).
When real-time data processing is required, it is preferable for the calculation units to be as close as possible to the data sources. This avoids the recurring problems of latency or unnecessary data overload encountered with more traditional Cloud solutions.
The use of edge computing / analysis favors local data processing at the level of edge gateways, or even at the level of the connected devices. Only data that cannot be processed locally, or that needs to be put online, is uploaded to the Cloud, which simplifies compliance with IT security requirements.
In addition, when locally generated data is high frequency (in the order of seconds to minutes) it can require very high bandwidth. In these cases of large volumes of data, it is usually not possible to transfer data in real time from the mainframe to the Cloud. This problem can be avoided by decentralized processing of this data at the edge of the network.
 AI: Artificial intelligence is the implementation of a number of techniques designed to allow machines to mimic a form of real intelligence.
 Edge computing, or computing at the edge of the network, is an optimization method used in cloud computing that consists of processing data at the edge of the network, close to the source of the data.
 EAM: Enterprise Asset Management. EAM refers to the management of an organization’s physical assets, such as buildings, facilities, infrastructure and other equipment, over their entire lifecycle.
 CMMS: Computer-aided maintenance management is a software-assisted management method for a company’s maintenance departments to assist them in their activities.