Artificial Intelligence (AI) is already strongly present in the worlds of IT and finance. However, the industry needs good AI products, and the Ghent-based Yazzoom saw an opportunity in this area. As a specialist in AI, machine learning and data analysis, they developed their own software product for the industry: Yanomaly. And the results are impressive: Yazzoom invariably scores very high at hackathons, and Yanomaly has been in daily use by its first customers for more than 2 years now.


From project to product

Yazzoom was founded in 2011 when AI and machine learning were still unknown to many. “We started as a project company, but over the last 6 years we have also been researching and developing our own standard products,” says Alexis Piron, Sales and Marketing Manager at Yazzoom. “We have developed our own software for monitoring and analysing industrial and IoT data: Yanomaly. Such a product did not yet exist for the industry.”

Industrial applications

“Our customers are mainly process and manufacturing companies in a wide variety of sectors: from automotive to chemicals and energy to food,” says Alexis Piron. “We analyse data from production systems for applications such as energy optimisation, predictive maintenance, quality improvement and anomaly detection. We combine our expertise in AI with specific industry knowledge: our data scientists are engineers in physics, chemistry and mechatronics, among other fields.”

 “We combine our AI and machine learning expertise with specific industry knowledge.” – Alexis Piron

Hackathons

Yazzoom successfully demonstrated the effectiveness of its working method during the Asset Performance 4.0 virtual conference at which various hackathons were held. Yazzoom took part in two predictive maintenance challenges: for Fluvius electrical substations and Aquafin pumping stations. Yazzoom won the first hackathon and was runner-up in the second.

A hackathon is an event in which software and hardware developers develop a fully functional solution to a specific problem within a specific time frame. Companies often organise hackathons as a more informal form of an RFQ, putting the products and expertise of suppliers to the test.

Focus on a single specialisation

“Our products regularly achieve the best results in objective benchmarking,” says Alexis Piron. "This is down to our narrow focus on industrial AI applications. You cannot be the best in every domain, but you can be the best in a single specialisation. We want to continuously deepen that expertise. Four years ago, we already had top algorithms for anomaly detection and failure prediction. After that, we built a graphical user interface around them that allows customers to perform their own data analyses.”

“Our products regularly achieve the best results in objective benchmarking. This is down to our narrow focus on industrial AI applications.” – Alexis Piron

Partnerships

“We supplement this expertise by setting up partnerships with other domain experts”, concludes Alexis Piron. “For example, we ran the hackathon at Fluvius together with our client Engie-Laborelec. They have very specific knowledge of electrical installations, we of data analysis and AI. In this way we bring together the right expertise for each customer and each sector.”

Would you like to hear some testimonials from Belgian companies that implemented interesting real-world AI applications together with Yazzoom and others? Then dive into chapter 5 of our free online course 'AI in Business', where you will find 35 actual use cases.