A whole 365 days ago, four researchers set out to encourage the industrial sector to make use of machine learning tools in decision-making. To do so, Gregory Nain, Asaad Moawad, Thomas Hartmann, and François Fouquet launched DataThings, a company based out of the Paul Wurth InCub incubator.
Over the past year, the startup has seen considerable evolution. Most importantly, DataThings created a premiere data analytics program, GreyCat. The program stores and treats large-scale, real-time data, such as sensor interpretive information, and in turn provides managers with acute decision-making knowledge. With the invent of live learning, this technology will likely see further innovation.
GreyCat has enormous potential. The DataThings team aims to demonstrate how large-scale the program can be—indeed, its dynamic nature allows use across many domains in the industrial sector.
DataThings gave Silicon Luxembourg a direct example. “In the domain of steel industry,” says the team, “it is very challenging to predict the outcome of processes and ensure a constant level of quality, mainly because sensor equipment and monitoring solutions reach physical limitations due to the harsh conditions of the blast furnace. To bridge the gap, Paul Wurth engineers put together years of know-how and physical and chemical models.
“The DataThings team aims to demonstrate how large-scale the program can be—indeed, its dynamic nature allows use across many domains in the industrial sector.”
“DataThings has been working closely with Paul Wurth process engineers to deploy deep learning algorithms on production history, with the aim of creating a production model for each process. This machine-learnt model combined with process know-how provide improved quality predictions and allow a better steering of the blast furnace process.” Paul Wurth has extended its solid relationship with DataThings at the same time that the startup obtained the “Made in Luxembourg” label.
But DataThings is working outside of the steel industry as well, lending credit to their claim of cross-industry application. They have collaborated with Creos, Cyprès international, and the young Parisian startup GISGO. A contract with the Lyon Metropolis aimed at developing the AI for a water resource management system as part of the European H2020 bioTope project is also underway.
“The objective for 2018 is to continue to make known DataThings and GreyCat, thanks to new research and development’s partnerships in the industrial sector but also to increase the team of the young company,” commented the DataThings team.