A two-step machining and active learning approach for right-first-time robotic countersinking through in-process error compensation and prediction of depth of cuts

Leco, Mateo; McLeay, Thomas E.; Kadirkamanathan, Visakan

doi:10.1016/j.rcim.2022.102345

Cited by 11 publications

(1 citation statement)

References 34 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The available techniques rely on algorithms that actively selects informative samples to be labeled by the human expert. An acquisition function is used to score the unannotated data which will then contribute to a faster learning of the machine learning model [48]. (2) To avoid the extensive testing for optimal hyperparameters, recent work has proposed the differentiable spectrogram where the window size and hop length are jointly optimized with the model parameters [49].…”

Section: Signal Processing For Machining Optimizationmentioning

confidence: 99%

A Concept from Sensor to Sustainability in Machining – An Interdisciplinary Approach over a Wide TRL Range

Ståhl,

Mikkelsen,

Lenrick

et al. 2024

Advances in Transdisciplinary Engineering

View full text Add to dashboard Cite

This publication describes a concept that intends to enable an optimization of machining with regard to the balance between criteria related to technology, economy and sustainability. The work is of a discussion nature and intends to provide a framework for further research and development in the area. Previous research and development during the 80’s and 90’s is presented in general terms and in particular the reasons for its limited success in providing real-time feedback on machining operations are highlighted, despite very large financial investments even by today’s standards. Ongoing research worldwide in current process optimization and its associated building blocks will be highlighted, and identified important work is referenced. Below are new conditions that can be linked to both process knowledge and its modeling, as well as new conditions for developing integrated sensors that can handle the extreme environment in and around a processing operation. A previous limiting factor has been signal processing and signal transmission, which with new knowledge and developed technology in the last 10 years provides new conditions for process optimization in real time.The need for new and up-to-date principles for process optimization, which also integrate sustainability issues and environmental impact, has increased in importance in several respects. Important issues such as tool utilization, efficient use of materials and high time utilization have become relevant as these process results control both energy consumption and environmental impact. The geopolitical development linked to the availability of critical tool materials such as cobalt and tungsten also drives research issues that can generally optimize and streamline production processes.Finally, the publication describes the possibility of realizing a real-time feedback and optimized machining that takes into account technology, economy and sustainability, through interdisciplinary research across several levels of technology readiness (TRL). The results are expected to have positive effects on several production factors before, during and after machining. Developed technologies in machining can also make valuable contributions to the development of other products and production processes.

show abstract

Section: Signal Processing For Machining Optimizationmentioning

confidence: 99%