2021 Open Innovation Factory Winners

How to improve the safety, reliability and time spent on inspecting GRTgaz’s network?

Currently, due the lack of efficient ATEX*-certified endoscopes on the market, gas-containing pipelines are emptied every time a new inspection is performed. IT Concepts and Induscope worked together on this Call for Projects to come up with complementary solutions that won over the jury made up of professional experts and operations managers. Both companies were named as winners.

* ATEX: European standard for explosion-proof devices.

IT Concepts, a German company with 17 years specialist expertise in industrial endoscopy, proposed a specially designed videoscope. “We had developed an ATEX zone 2* videoscope for an American gas client. We then began a research phase to modify the equipment to GRTgaz’s specifications and come up with a perfectly insulated, air-tight solution,” explains Nicolas Rossi, Head of French Sales for IT Concepts GmbH.

* European standard. This aligns with “Level 3” under German law.

Induscope is a French company specialising in industry endoscopy. It provides endoscopes and pipeline inspection services for a wide range of industrial sectors (aeronautics, automobile, wind power, pharmaceutics). IT Concepts’ proposal is supplemented with a cylinder seal installed on the “tapping points” (intersections between the cylinders). “We developed a cylinder seal to seal these tapping points during the inspection process,” explains Guillaume Conné, Induscope Managing Partner. “The operator now only has to position the cylinder seal to render the pipelines inert and introduce the video-endoscope.” The inspection can be performed safely without having to empty the pipelines, saving precious time.

GRTgaz is currently studying these proposals to confirm that they meet the needs and safety standards for industrial sites and facilities.

DAVAI!JPL is an incubator for hi-tech projects, connected objects, and cutting-edge industrial technologies. The company's pneumatic robot-tube, Ivy Spec, is able to sneak into confined spaces, such as pipelines, and adjust itself to their diameter, for which it received the Special Mention jury award. The robot-tube is steered by a central AI control. With no stiff metallic parts, it can navigate independently and position its radars within the tightest nooks. “There are numerous applications, from aircraft fuel tanks to caving rescue operations,” adds Jean-Philippe Lerat, President of Davaï!JPL.

How to optimise GRTgaz’s industrial assets with the best possible exploitation of reliability and maintenance data?

“This was a topic that resonated strongly with us due to the intersection of our many assignments to optimise maintenance schedules for industrial-scale Energy and Transport companies,” says Paul-Alexandre Bouhana, Data Science Project Manager at Artelys, an independent SME specialising in data science, modelling, network optimisation and regional energy planning. Artelys proposed AI algorithms and graphics to make the results comprehensible and provide a better understanding of failure probabilities. “We also came up with a supplementary decision-assistance module. By identifying how imminently a particular part was liable to fail, and its impact on the network, we could prioritise maintenance actions and adjust work schedules. The fact that we went above and beyond the initial request was what impressed the jury.” Artelys’ first task will be to develop a prototype to test the project’s viability from the first half-year for GRTgaz.

Sia Partners, a management consultancy firm, has developed a specific expertise in energy sector data science.

Romain Guichard, Senior Data Science Consultant, headed up the proposals. The concept involved a predictive model for asset failures. How was this done? “By cross-referencing each item of equipment’s history and the corrective actions undertaken in order to identify the predictor variables.” Knowing both the probabilities and the causes of a malfunction means we can make accurate predictions linked to the equipment’s usage conditions. “We have demonstrated our model’s accuracy on pilot sites, and we are already developing tools for its industrialisation,” concludes Pierre Leplatois, European Data Science Business Unit Director.