The CAPRI project reaches its conclusion
Authors: Konstantina Tsioli, Ioannis Maimaris, Ilia Kantartzi
7th November 2023
After four years, the CAPRI project has come to an end with its final results and an online closing conference. CAPRI’s main goal was to develop cognitive solutions to the Process Industry, in order to facilitate its Digital Transformation. CORE Innovation Centre has been involved in various tasks as part of the project, and we are excited to see project outcomes reach their maturity.
As part of the project, our team developed advanced deep learning models for anomaly detection and Remaining Useful Life estimation of critical components in the asphalt use case. We were also responsible for delivering some of the consortium’s exploitation activities, as well as fully managing all project-related communication tasks.
The goal of CAPRI was to develop, test, experiment and deliver an innovative Cognitive Automation Platform which incorporates cognitive technologies, such as artificial intelligence, machine learning, and advanced automation, to enhance the operations within the Process Industry. Critical outcomes include actions to enhance the flexibility of operations, making the processes more adaptive and responsive, as well as actions to improve operational performance by reducing costs, improving maintenance efficiency, optimising resource utilisation and more.
Deep learning models
The deep learning tools developed by our team have been applied to a use case for asphalt production (EIFFAGE), aiming to reduce maintenance and spare parts costs related to critical operations and to enhance the reliability and robustness of the maintenance system. More specifically:
Anomaly Detection: By leveraging deep learning techniques, this model excels at identifying potential malfunctions in the machinery of asphalt use case. It acts as an ever-watchful guardian, constantly monitoring the baghouse system to alert for anomalies before they become critical issues. This proactive approach allows maintenance to be optimised and to minimise unexpected disruptions.
Remaining Useful Life (RUL) Estimation: Extending the anomaly detection model, our team went one step further by estimating the remaining useful life of critical components. In the EIFFAGE use case, the critical sensor is located at the entrance of the baghouse. This component is essential for maintaining the efficiency and quality of the involved processes. As an outcome, through our RUL estimations, we can accurately predict the time until the next failure of this critical component, offering manufacturers with the foresight needed to plan maintenance activities effectively.
CORE Anomaly Detection model for critical component constant monitoring and for providing possible alerts prior malfunctions.
These solutions have implications for other industries and, once applied, can potentially increase cost efficiency in the steel, aluminum and copper, cement, pharmaceuticals and glass manufacturing industries. More information on the EIFFAGE use case can be found here.
The commercialisation phase
Our innovation team contributed to the exploitation of CAPRI project outcomes by analysing the financial sustainability of the applicable exploitable results, utilising our custom Profit Simulation Tool. This endeavor aimed to gain insights into the financial requirements and resources necessary to introduce the solutions to the market and identify feasible scenarios for the commercialization phase.
This involved estimating Revenues and Costs for a 5-year post-project horizon. The knowledge accumulated throughout the project, which involved the analysis of market conditions and customer segments, was further developed, projecting this analysis into the future for the market. Initially, the analysis focused on customer segments related to the project use case industries, with plans to later expand to include industries identified through the replication analysis. Various scenarios were examined, to pinpoint a pragmatic and viable strategy for partners to implement so they can successfully bring their solutions to market and deliver a sizeable impact for CAPRI on the EU process industry, while also developing their business.
Communication activities
On the communication side, our consortium participated in 32 events over the years, with a total of 13 publications and 16 articles published. The project performed exceptionally well on social media, garnering over 900 followers on X (formerly known as Twitter) and over 1300 on LinkedIn. Additionally, our team ran a YouTube account, which hosts 21 videos with over 2000 views in total. The project website, which was designed by CORE IC and officially launched in the early months of the project, has gained 8400 visits in the three-year run of the project, and it will continue to serve as a central hub for all project deliverables.
