With the advent of Industry 4.0 and Smart Manufacturing paradigms, data has become a valuable resource, and very often an asset, for every manufacturing company. Data from the market, from machines, from warehouses and many other sources are now cheaper than ever to be collected and stored. A study from Juniper Research has identified industrial internet of things (IIoT) as a key growth market over the next five years, accounting for an increase in the global number of IIoT connections from 17.7 billion in 2020 to 36.8 billion in 2025, representing an overall growth rate of 107%. With such an amount of data produced every second, classical data analysis approaches are not useful and only automated learning methods can be applied to produce value, a market estimated in more than 200B$ worldwide. Using machine learning techniques manufacturers can exploit data to significantly impact their bottom line by greatly improving production efficiency, product quality, and employee safety.
The introduction of ML to industry has many benefits that can result in advantages well beyond efficiency improvements, opening doors to new opportunities for both practitioners and researchers. Some direct applications of ML in manufacturing include predictive maintenance, supply chain management, logistics, quality control, human-robot interaction, process monitoring, anomaly detection and root cause analysis to name a few.
This workshop will ground on the successful story of the first edition, with 19 oral presentations and 3 invited talks, to draw attention to the importance of integrating ML technologies and ML-based solutions into the manufacturing domain, while addressing the challenges and barriers to meet the specific needs of this sector. Workshop participants will have the chance to discuss:
● needs and barriers for ML in manufacturing
● state-of-the-art in ML applications to manufacturing
● future research opportunities in this domain
Call for Papers
This is an open call for papers, soliciting original contributions considering recent findings in theory, methodologies, and applications in the field of industrial machine learning. Position papers presenting industrial use cases and discussing potential solutions are welcome. Potential topics include, but are not limited to:
Robustness-oriented learning algorithms
Machine learning for robotics (e.g. learning from demonstration)
Continuous and life-long learning for industrial applications
Transfer learning and domain adaptation
Anomaly detection and process monitoring
ML applications to Predictive Maintenance
ML applications to Supply Chain and Logistics
ML applications to Quality Control
ML for flexible manufacturing
Deep Learning for industrial applications
Learning from Big-Data
Inference in real-time applications
Machine Learning on Embedded and Edge computing hardware
All the contributions are expected to expose applications to the industrial sector, possibly with real world case studies. Position papers presenting new industrial systems and case studies, possibly reporting preliminary validation studies, are also encouraged.
A Best Paper Award will be assigned to the most relevant contribution. The decision will be taken by the Organizing Committee according to the reviewers’ feedback.
Submission
Papers must be prepared according to the ICPR guidelines (see here). All papers will be reviewed by at least two reviewers with single-blind peer review policy. Papers will be selected based on relevance, significance and novelty of results, technical merit, and clarity of presentation. Papers will be published in ICPR proceedings.
Templates in Word or Latex can be downloaded here: word | latex
All the papers must be submitted using CMT submission server.
Important dates
Full Paper Submission:
May 7, 2022EXTENDED to June 10, 2022Notification of Acceptance: June 20, 2022
Camera-Ready Paper Due: September 1, 2022
In case of rejection from ICPR main conference, authors can submit their work to the IML workshop. Authors should address all ICPR reviewers' comments in the submitted paper and submit the ICPR reviews as supplementary material.
Program (Tentative)
9.00 AM - Welcome & Opening
9.10 AM - Invited talk: Martin Atzmüller (Osnabrück University and German Research Center for AI, Germany) - Explainable & Interpretable Machine Learning: Methods and Applications in Complex Industrial Contexts
10.00 AM - Coffee break
10.30 AM - Geri Skenderi: "On the use of learning-based forecasting methods for ameliorating fashion business processes: A position paper"
10.45 AM - Buddhika Laknath Semage: "Intuitive Physics Guided Exploration for Sample Efficient Sim2real Transfer"
11.00 AM - Invited talk: Marco Rudolph (Leibniz University Hannover, Germany) - Industrial Anomaly Detection with Normalizing Flows
11.30 AM - Invited talk: Danijel Skočaj (University of Ljubljana, Slovenia) - Data-driven learning-based surface defect detection
12.10 PM - Saeid Rezaei: "Gym-DC: Reinforcement Learning Environment in Distribution Centre"
12.25 PM - Marco Cristani: "Toward Smart Doors: A Position Paper"
12.40 PM - Francesco Setti: "IndRAD: A Benchmark for Anomaly Detection on Industrial Robots"
Organizers
Francesco Setti
University of Verona
Luigi Di Stefano
University of Bologna
Paolo Rota
University of Trento
Vittorio Murino
University of Verona
Massimiliano Mancini
University of Tübingen