AI is now everywhere: in homes, with intelligent conversational agents, in cities, with the introduction of autonomous cars on a trial basis, on the Internet with browsing that considers your preferences, but also in companies…
Xavier Brucker
Partner | Polytechnique MIT
Christophe Bressange
Senior Partner | EMLyon
Jong-Mo Allegraud
Senior Data Scientist | ISAE
“AI is a set of methodologies particularly adapted to the world of industry. It is important to invest early in AI, to enrich internal skills and generate a strong competitive advantage.”
Xavier Brucker
Partner | Polytechnique MIT
A graduate of Ecole Polytechnique (X 1996), Telecom ParisTech (2001) and the Massachusetts Institute of Technology (MIT 2002), Xavier combines several international experiences with industry companies and start-ups. After a career in strategy consulting, he led a major defense program at Safran and then became director of the e-commerce and payments division of EquensWorldline. Xavier then spent several years in California, in the heart of start-up companies in Los Angeles and Silicon Valley. Upon his return to France, Xavier joined a Fintech start-up, which he supported in its growth. Today, within Mews Partners, he develops offers related to data science and artificial intelligence, which have become a real strategic challenge for industry and services.
Christophe Bressange
Senior Partner | EMLyon
Christophe is a graduate of EM Lyon and has more than 15 years of experience in leading supply chain consulting firms. He also held operational management positions for 7 years in the Geodis group and the Casino group. Christophe joined the Operations practice to bring his knowledge of the GIC, Retail, Transport & Logistics sectors, as well as his experience in international supply chain transformation projects.
The Artificial Intelligence (AI) and Data Science approach is particularly relevant in industrial contexts.
AI will monitor, predict, and optimize predefined KPIs and provide real-time answers. The complexity related to many different parameters will be easily manageable by a machine learning model. Man will still have added value in interpreting context and exceptions, and thus making the right decisions with the help of the machine.
An AI project consists in properly framing the subject and the KPIs that need to be optimized and in anticipating the interaction of these methodologies in the global process as well as with the people who lead it.
of production data is not used
of company data is not used
zettabytes of data in the world by 2020
In organizations, Artificial Intelligence will first seek to improve operational efficiency. In practice, the most effective AI approaches are based on the analysis of KPIs related to the processes studied and seek to predict and optimize them on the basis of a large number of parameters. Products, services, processes and life cycle management can gradually be automated under the control of human operators. The company then transforms itself relying on predictive and prescriptive functionalities, capitalizing on past events. Robots, cobots, natural language and image processing, but also RPAs (robotic process automations) are among the methods mastered and deployed.
It is essential to start with a business vision of the problem you are working on. What are the KPIs? What are the parameters that can influence these KPIs? What types of results would be useful?
The actual “data science” phase consists of gathering data, cleaning it if necessary, and injecting it into different models to find the best performance.
The last phase is crucial: it is the insertion of the model into the industrial process, taking into account the added value of human supervision, the performance and reliability of algorithms.
AI is particularly suitable for industrial projects.
Are you interested in these topics?