Periodic Reporting for period 3 - El-Peacetolero (Embedded Electronic solutions for Polymer Innovative Scanning Tools using Light Emitting devices for diagnostic Routines)
Reporting period: 2023-09-01 to 2025-02-28
A system performing on site non-destructive identification of polymers and diagnosis of the degradation of these material is needed to implement the new safety requirements of the European nuclear industry. El-Peacetolero overarching ambition is to design a TRL 7 hand-held, low power, embedded optoelectronic system that can deploy AI for in-situ real-time measurement, identification and degradation diagnosis of polymers in an industrial environment.
WP2: Measured data were labelled and uploaded to the exchange server. Algorithms for simulation and statistical augmentation were developed. Machine Learning polymer identification model concept was developed. It enables, with a hierarchical decision, more accurate prediction of material degradation. The models were successfully embedded in the final El-Peacetolero Tiny ML prototype device. The execution times were less than a millisecond, which is considered real time. The memory consumption is very low, in the order of Kb.
WP3: The new modular architecture of El-Peacetolero battery powered prototype based on “Tiny ML platform” was designed and realized. Tiny ML platform is the system’s Mastercard, it provides sufficient performances to run AI algorithms, EDGE computing and Tiny ML. The Analog bloc PCB and the DAS PCB performances was improved, their dimensions divided by 3 and power consumption reduced by 70%. A demonstrator of the final prototype of El-Peacetolero Tiny ML embedded optoelectronic system has been designed and experimentally validated. It is compact, lightweight, ergonomic and can be held by a technician or handled by robot to perform measurement on polymers. Its measurement accuracy is lower than 1%. It was validated by carrying out measurement on aged Neoprene and HDPE samples. Prototype measurements at 3.4 µm are consistent with sample’s FTIR-ATR characteristics. AI algorithms implemented in El-Peacetolero Tiny ML embedded optoelectronic system prototype performs real time polymer identification and ageing diagnosis with low memory consumption.
WP4: Compact multi-lambda QCL IR source was developed and validated. Device dimensions was divided by 50. It was used with IR detector, ATR crystal and control electronics to build EL-PEACETOLERO laser-based measurement head prototype.
WP5: The LED based measurement head was developed using an appropriate photodetector and IR LED emitting at the 3.4 µm selected wavelength. Device relative measurement uncertainty is lower than 1%. The measurement head is compact and ergonomic.
WP6: System hardening : Two irradiation tests were performed by carrying out pre-irradiation, online, and post irradiation tests. Most of the electronic parts as well as the prototype remain functional at doses above 100 Gy, the required dose for the hardened version of the system.
Material testing with El-Peacetolero prototype: measurements on Neoprene and PE samples was carried out with Laser based and the LED based measurement heads. Prototype’s measurements stability confirm that it has a TRL 6 and it will achieve a TRL7 with a revised design reducing ATR crystal size. Measurements with the LED based prototype were carried out on several Neoprene and PE100 samples issued from accelerated thermal ageing experience at T=80°C. Measurement results are consistent with ATR-FTIR. El-Peacetolero LED based prototype performs polymer identification and ageing diagnosis. It has at least TRL 6 and it will achieve TRL7 by carrying out measurements in industrial environment.
Terahertz measurements show that polymer aged samples can be distinguished from each other if the ageing conditions are known. However, it is not possible to distinguish the samples with unknown ageing conditions.
FZJ has conducted a remarkable in-depth study to determine the aging process of Neoprene by the physico-chemical processes occurring in the material. They demonstrate that aging process of neoprene can be divided into three stages. FZJ has done an outstanding work in proposing a mechanical characterisation of the samples, which made it possible to establish a correlation between FTIR-ATR measurements and the degradation processes.
Elongation at break (EAB) was performed on 45 rectangular pieces of Neoprene, PE and PERC aged under different ageing conditions. EAB for Thermal Neoprene aged at 80°C and for immersion Neoprene follow a linear change.
WP7: Three prototypes of the robot have been constructed: Prototype 1, Prototype 2 100 % and Prototype 3 have been completed with an underwater electronics cage. Prototype 3 features a modular mobility system, allowing it to be configured with tracks, thin wheels, or omnidirectional wheels depending on the surface conditions inside the pipeline. The final stage of the project involved the public demonstration of all three robotic prototypes at Sorbonne University in February 2025. This event showcased the full capabilities of Prototype 3, as well as the mechanical integration of the Peacetolero sensor into Prototype 2, marking a significant milestone in underwater pipeline inspection technology.
A set of aged neoprene, PE100, and PE100RC polymer samples been and their FTIR-ATR measurements had produced in the project. Analysis of the produced FTIR-ATR data and mechanical test afforded the project team to correlate the FTIR-ATR characteristics to the ageing state of the polymer. The project team established an ageing threshold to be used in the El-Peacetolero prototype do alert to ageing conditions of neoprene and PE100 polymers. This key result is of a major importance for polymer diagnosis.
The second Key result is related to the development of multi-lambda QCL source which integrate 4 QCL laser in a compact PCB. With El-Peacetolero results it is possible to have more than 15 QCL lasers a compact measurement head. The third Key result is the development of the Tiny-ML Elpeacetolero embedded optoelectronic system. This compact intelligent sensor affords Edge computing and in-situ Machine Learnin. Further, three robot prototypes have been developed in the project.
The fourth key results is the robot prototypes demonstrating the feasibility of deploying modular robotic systems for pipeline inspection and underwater intervention.
These results will have a potential impact in many other fields such as medical, environmental and aerospace applications. The system can be used for Point-of-care test (POCT), material control in aero crafts and deep-sea intervention.