Digital Twin-Driven Adaptive Sim-to-Real Alignment via Reinforcement Learning for Vibration-Based Bearing Health Monitoring Under Data Scarcity
arXiv:2606.24954v1 Announce Type: new Abstract: Vibration-based health monitoring of rotating machinery requires reliable fault diagnosis under operational data constraints, yet condition assessment remains challenged by structural scarcity of fault events and heterogeneous sim-to-real gaps in digital twin-generated signals. Each fault type generates impulses with distinct periodicity, amplitude modulation, and spectral character, making feature-space discrepancies fundamentally heterogeneous across fault classes. Existing domain adaptation methods apply a class-agnostic global transformation
The increasing complexity and automation of industrial systems necessitate more robust and data-efficient monitoring solutions, especially with advancements in reinforcement learning and digital twin technologies.
This development addresses a critical challenge in industrial maintenance by making predictive health monitoring more reliable and adaptable even with limited fault data, potentially reducing downtime and operational costs.
The ability to accurately diagnose machinery faults with sparser data and more accurately bridge the 'sim-to-real' gap enhances the scalability and effectiveness of AI-driven predictive maintenance across various industries.
- · Industrial IoT companies
- · Heavy manufacturing
- · Predictive maintenance software vendors
- · Critical infrastructure operators
- · Traditional reactive maintenance services
- · Companies reliant on extensive fault data sets
More accurate and automated early detection of machinery failures, leading to reduced maintenance costs and unplanned outages.
Increased operational efficiency and uptime across industries heavily reliant on rotating machinery, creating competitive advantages for adopters.
The proliferation of more resilient and autonomous industrial systems, paving the way for fully self-optimizing factories and infrastructure.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG