News About Clawdite

• Skill Mapping in the Circular TwAIn Project
• Clawdite Joins XR5.0
• Clawdite will be presented at CIRP-CMS2022
• Selecting Wearable Devices using AHP
• Clawdite published in Procedia CIRP

Skill Mapping in the Circular TwAIn Project

The Circular TwAIn project is redefining workforce optimization by integrating AI technologies into the industrial ecosystem. At the heart of this transformation lies Clawdite, our versatile platform, which plays a pivotal role in the Human Digital Twin (HDT) component. One of the project’s key achievements is the successful mapping of operator skills to the O*NET Content Model, enabling smarter and more efficient workforce management through AI-powered task assignment.

Mapping Operator Skills with O*NET

To model and standardize operator skills, we leveraged the O*NET Content Model (O*NET Content Model), a comprehensive taxonomy designed by the U.S. Department of Labor. This model serves as a structured foundation to describe work and worker characteristics in a consistent and analyzable format.

In Clawdite, the mapping process employed AbstractDescriptors to associate high-level worker data with O*NET taxonomy items. This design enables us to:

• Model skills as characteristics,
• Link each skill to a corresponding taxonomy item, and
• Connect each taxonomy item to a known Taxonomy instance, representing the O*NET model.

By structuring data this way, Clawdite makes it straightforward to retrieve an operator’s skills while filtering out unrelated characteristics (like age or height) stored in the HDT.

An Enriched Data Model for Clawdite

To support this advanced skill mapping, Clawdite’s data model was extended to:

1. Represent skills and values in alignment with a given taxonomy (O*NET),
2. Track task assignments and include a mechanism to model exemptions—tasks that an operator is explicitly prevented from performing.

Tasks, or interventions, are considered assignments issued by an orchestrator (e.g., a manager) to a factory entity (a worker). The new Exemption concept allows us to annotate specific cases where an operator cannot perform a task, further enhancing the precision of task assignment.

Introducing the Operator2Task Assignment Module

Building on this skill mapping infrastructure, within the Circular TwAIn project our lab developed the Operator2Task Assignment module. This AI-driven component intelligently matches workers to tasks based on skill compatibility, making it a cornerstone for efficient workforce orchestration.

Key Capabilities

• Skill Profiling: The HDT continuously records operator skills and task history, creating a comprehensive profile.
• Task Analysis: When task requirements are not predefined, the system uses AI (language models) to analyze task descriptions and infer needed skills via sentence embeddings.
• Vector-Based Matching: Both operator profiles and task descriptions are converted into vectors. Using similarity metrics (e.g., cosine similarity), the system computes a compatibility score.
• AI-Powered Recommendation Engine: The module suggests the best-matching worker for each task, along with a confidence score to aid decision-making.

This approach ensures not only that the most suitable operators are assigned to each task but also promotes higher job satisfaction by aligning tasks with workers’ strengths.

Conclusion

By enriching the Clawdite platform with skill mapping aligned to the O*NET Content Model and integrating AI for intelligent task assignments, the Circular TwAIn project is paving the way for next-generation workforce management. The Human Digital Twin becomes not just a data repository, but an active, intelligent system that transforms how organizations allocate human resources—optimizing efficiency, ensuring worker satisfaction, and embracing the power of AI in industry.

For more information about the project, visit the official site: circular-twain-project.eu

This work has received funding from the Swiss State Secretariat for Education, Research and Innovation (SERI) and has been partly supported by the European Union’s research and innovation programme under project Circular TwAIn (Grant n. 101058585).

Clawdite Joins XR5.0

Clawdite is now an integral component of the newly launched XR5.0 project, a forward-looking European initiative that merges cutting-edge XR technology, AI, and Human Digital Twins to redefine how people and machines collaborate in Industry 5.0 environments.

What Is XR5.0?

XR5.0 aims to revolutionize industrial extended reality (XR) by combining AI, human-centric digital twins, and European-made XR technologies to support a wide range of Industry 5.0 applications. These include:

• Ergonomic and personalized training,
• Intelligent asset maintenance,
• Worker guidance and monitoring,
• Human-AI collaboration via explainable and generative AI.

XR5.0 will implement and test six high-TRL pilot applications across real manufacturing environments, creating a strong foundation for sustainable industrial XR adoption in Europe.

Clawdite’s Role in XR5.0

To support the ambitious goals of XR5.0, we’ve deployed a dedicated shared instance of Clawdite. This instance acts as the central Human Digital Twin (HDT) knowledge base, storing structured data about workers and devices that feed into functional XR modules.

Key Use Cases Supported by Clawdite:

• Worker Movements Prediction:
  Clawdite stores and structures real-time and historical data on operator positions, enabling predictive models to monitor and forecast worker movements in shared spaces (e.g., environments with Autonomous Mobile Robots (AMRs)). This enhances safety and collaborative efficiency.

• Worker Shadowing & Monitoring:
  Leveraging Clawdite’s ability to represent skills, tasks, and device usage, this module supports the supervision and real-time assistance of less experienced technicians. The HDT data ensures that support is context-aware and personalized, aligning with XR5.0’s person-centric philosophy.

Why Clawdite?

Clawdite brings to XR5.0 a mature, modular platform designed to support Human Digital Twins (HDTs) in complex, real-world industrial scenarios. Building on the enhancements introduced in previous versions, Clawdite offers:

• A robust worker modeling framework, allowing precise representation of worker capabilities, assignments, and interactions.
• Real-time and historical data management, also ensuring continuity even in offline or disconnected environments.
• Flexible data access via HDM and HDM-Web, enabling retrieval of time-series and batch data.
• Unified and developer-friendly APIs, documented with OpenAPI 3 and accessible through a centralized Swagger UI, simplifying integration with XR functional modules.

With these capabilities, Clawdite serves as a core enabler of intelligent and person-centric XR applications, providing the intelligence layer needed to personalize, monitor, and optimize human-machine interaction in Industry 5.0.

This work has received funding from the Swiss State Secretariat for Education, Research and Innovation (SERI) and has been partly supported by the European Union’s research and innovation programme under project XR5.0 (Grant n. 101135209).

Clawdite will be presented at CIRP-CMS2022

Our paper on Clawdite will be presented at CMS2022, the 55th edition of the CIRP International Conference on Manufacturing Systems. The conference is organised by the University of Applied Sciences and Arts of Southern Switzerland - Department of Innovative Technologies and it will be held in Lugano between the 28th of June and the 1st of July.

Selecting Wearable Devices using AHP

Adopting wearable devices is crucial in Industry 5.0 applications, where humans become part of the digital representation of the factory, the so called Human Digital Twins. However, the selection of the devices to use turns to be cumbersome for practitioners and researchers, because of the wide availability of devices in the market.

We faced this criticality while adopting Clawdite in a real-world application, where the aim was to monitor worker’s physiological data so that to enhance the human-in-the-loop control of a production system.

We thus defined a methodology to support the devices selection, focusing on wearable devices for Industry 5.0 applications. The methodology is based on the Analytic Hierarchy Process method, and it allows us to successfully identifying the two devices that satisfy our needs the most, among a list of 110 alternatives.

Here below we report the list of the 89 devices we ranked using our methodology (we discarded obsolete devices and those not fitting the application requirements from the initial list of 100 devices). The top-20% devices are highlighted.

Wearable Classification

Clawdite published in Procedia CIRP

In the context of the Industry 4.0 approach, applications and solutions supporting monitoring, simulation, optimisation, and decision-making in production systems are exponentially growing.

Digital twins are the silver bullet for these solutions, given their capability of providing comprehensive digital representations of production systems, which are continuously updated by leveraging IIoT (Industrial Internet of Things) systems.

However, digital twins designed for Industry 4.0 applications are limited in representing human operators, which instead are a key component in the emerging Industry 5.0 paradigm.

Cognitive automation is spreading increasingly, and it becomes essential to have a digital representation of humans for improving their well-being and performance, but also to monitor and simulate their interactions with the factory.

A standardised solution for creating digital twins is currently missing. Industrial solution architects are forced to resort to ad-hoc implementations and models, which lack re-usability, scalability and extensibility. These limitations prevent a smooth introduction of a human digital representation in existing digital twins, so hindering the complete shift to the new Industry 5.0 paradigm.

The SPS Lab worked on a solution to bridge the existing gaps with current digital twins, by introducing Clawdite, an extensible and flexible IIoT-based platform for creating Human Digital Twins. Clawdite supports the creation of customised data representations of production systems, including machines, devices, and humans. Moreover, Clawdite has been designed to be modular, so that to ease the integration of new components (e.g., for simulations), the digital twin instantiation, and the system ramp-up.

Clawdite has been tested in different projects (STAR, SCOUT), showing its capability to fit different applications and laboratory settings. The resource is publicly available in GitLab.

Want to know more about Clawdite, its architecture and flexible data model? Checkout our paper! If you have any questions, contact us (project key people: Andrea Bettoni, Elias Montini).