ICSA 2025
Mon 31 March - Fri 4 April 2025 Odense, Denmark

This paper introduces a scalable and dynamic architecture for AI pipelines designed to address challenges in distributed systems and industrial AI applications. The architecture employs a modular, service-oriented structure integrated with the Heartbeat-Update-Synchronize-Ingest-and-Register (HUSIR) protocol to enable efficient task orchestration, fault tolerance, and dynamic routing of jobs. Automated workflows replace manual data processing, significantly improving scalability, adaptability, and efficiency. Mechanisms for edge-case data collection and non-disruptive real-time testing align AI development with production needs, fostering continuous model improvement. The proposed system was evaluated over two years in collaboration with Esoft, a global leader in visual content solutions for real estate. The evaluation demonstrates substantial improvements in processing speed, resource utilization, and production capacity, highlighting the architecture’s effectiveness in scaling automated production systems and handling diverse, dynamic workflows. This work provides valuable insights and practical solutions for implementing robust, adaptive AI pipelines in industrial settings.

Thu 3 Apr

Displayed time zone: Brussels, Copenhagen, Madrid, Paris change

16:00 - 17:00
Industrial IoT, Edge, and Cyber-Physical Systems IINew and Emerging Ideas / Research Papers / Software Architecture in Practice at Hall 2 (U82)
Chair(s): Adel Noureddine University of Pau and Pays de l'Adour
16:00
15m
Research paper
Non-invasive software architecture for data pipelines with legacy support in smart manufacturing
Research Papers
Alberto Ceselli , Giuseppe de Martino , Patrizia Scandurra University of Bergamo, Italy
16:15
15m
Paper
AI Pipelines: A Scalable Architecture for Dynamic Data Processing
Software Architecture in Practice
Jakob Hviid University of Southern Denmark, SDU Software Engineering, Anders Launer Bæk-Petersen University of Southern Denmark, SDU Software Engineering, Emil Stubbe Kolvig-Raun University of Southern Denmark, Universal Robots, Juan Marín-Vega
Pre-print
16:30
15m
Paper
Architectural Proposal for Reproducible, Standardized Deep Learning Research
Software Architecture in Practice
Max Luebbering , Vijul Shah , Moinam Chatterjee Otto-von-Guericke University Magdeburg, Germany, Priya Priya , Osama Soliman , Rafet Sifa
16:45
15m
Paper
Scalable Architecture for Intent Based Optimal Control of Composite Systems
New and Emerging Ideas