PRODE
OPEN CALL 1 | INNOVATION SUPPORT
Name: PRODE SMART PORTABLE DETECTOR
Flexible manufacturing systems (FMS) | Energy efficiency | Decision support systems
Abstract: Current Fabric Quality Detection executed by Human operators is not free of errors impacting on subsequent production machineries, waste end energy amount. It requires human resources with future probable cost increasing due to lack of personnel availabilities, vice-versa the implementation of fixed automated solution requires plant revolution with significant CAPEX expenditure and slow ROI.
PRODE – Product Detector is a proposal to innovate processes into the Textile Production Industries regarded fabric quality control. Based on cyber vision apply to Big Data, Analytics, Artificial Intelligence, Digital Twin, Cloud topics of Industry 4.0. Its features are: portability, easy to setup in the environment, flexibility with a centralized server side logic AI based. Prode is equipped with easy to read dashboard analytics for decision support system.
Objective of the proposed solution is introduce digitalization in fabric inspection process to make it faster, automated and regarded a deeper investigation then the human eye can provide reducing time, wastes but at same time supporting a collaborative human and not human industrial environment development with linear CAPEX expenditure in parallel with incremental-non disruptive approach to AI based automation.
Initial TRL was between 2-technology concept formulated and 3-experimental proof of concept. At the beginning of the project TR2 already had already observed the basic principles, formulated a technology concept and started a experimental proof of concept in order to validate the following initial hypothesis: “using Artificial Intelligence equipped with deep sensing and appropriate training we can be able to detect anomalies resulting from video detection”. After 3 months from the beginning of the project we achieved a TRL of 4, Technology validated in lab. At the end of the development the TRL is 7, we assembled a prototype tested in operational environment.
Innovation is regarded different aspects:
• From human eye to cybervision: visual inspection modules armed with Artificial intelligence executes the operator task in less than half time extending it to a next level of detection about minimal anomalies, structure and or pollutants.
• From fixed equipment to portable tools for a efficient use where needed: Prode is portable can be placed anywhere any moment when needed and executes the detection of the fabric. Prode portability assure zero impact on current production chain, no changes and no remodulation are needed.
• From zero Data to the anomalies catalogue: strategic decisions are supported via KPIs and analytics.
Users are the fabric quality control operators in Textile markets. Same approach can be extended to other sectors where raw materials are production assets.
Results of the experiment are: Tr2 developed a test bed to simulate the production environment according to the suggestions received from the Manufacturing Partner, Tr2 developed AI algorithms experimenting different approaches, applied the elaboration to live and recorded data provided by the Parter and/or generated in the test bed and collected the KPIs in a friendly operator dashboard.
The solution has been disseminated in different meetings with manufacturing top players like Roteks and AIpplas in Turkey and presented live in a meeting with the partner where tests have been executed.Tr2 presented the project KPIs achieved as planned and performed a next steps plan with the partner according to a next level of desiderata and a next level of business exploitation.
All results, software repositories, video link of the solutions have been published including the final pitch video with the experimentation presented in Izmir in the partner premises. A video shooting has been also produce by the M4M video team.
Main applicant
- Name: Tr2 srls
- Country: Italy
Industrial partner
- Name: MOSI Tekstil
- Country: Türkiye
