8.01 Human Digital Twin for future manufacturing systems

REFERENCE SPOKE
OTHER SPOKES
PROJECT LEADER
Daria Battini
PROPOSER
Università degli Studi di Padova
PARTNERS

Politecnico di Milano, Politecnico di Torino, Università degli Studi di Brescia, Leonardo S.p.A. 

8.01 Human Digital Twin for future manufacturing systems

If humans and automation systems must complement their capacities in order to achieve improved manufacturing performances, then the operator’s historical data, status and evolution must be available for analysis and decision-making. Thus, it becomes imperative to create digital representations not only of production systems and robots but also of workers, by considering context data, static data, real time sensor data, and natural language data. In one word, the Human Digital Twin (HDT) becomes central in the development of effective industrial collaborative applications. While Digital Twins (DT) started to spread in aerospace sector, Human Digital Twins were born from the health sector, where they were conceived as a sophisticated simulation of the human body. In the industrial work field, HDT can be implemented in a closed control loop, leading to rapid shop floor reconfiguration for optimized production quality performance and ergonomics purposes, without the need to interrupt the production process. This project aims to create a new proof of concept for human digital twin in manufacturing in order to design the human-centric workplace of the future that will be inclusive, sustainable and resilient. The researchers involved in this project will develop a new digital representation of the workers by collecting real-time data regarding workers’ status, well-being, health/safety parameters (including postures and fatigue) and performance evolution in order to support the operation manager decision making. A multi-sensors data fusion will be investigated and integrated with environmental data and workers’ historic data. Digital ergonomics tools, biosensors and wearable sensors, Extended Reality (XR) technologies coupled with Machine Learning (ML) and Natural Language Processing techniques (NLP) are strategic in this context to finally design a digital twin for human worker and create effective and sustainable collaborative working environments. The research activities will at first develop a framework for the conceptualization of collaborative human-machine working environments through physical/physiological tracking, behavioral analysis and mixed reality technologies using also conversational Artificial Intelligence (AI) algorithms and systems. In a second phase we will provide technical specifications and baselines for developing human digital twins in manufacturing, enabling the design and development of future sustainable work environments also by assisting workers’ tasks with Augmented Humans solutions. This concept will permit us to realize socio-technical digital twins and making a step forward the current industrial digital twins.

RISULTATI ATTESI

The main result of this project will be the proof of feasibility of human digital twin realization and implementation in a real manufacturing context. A real-time monitoring by a human digital twin will prevent prolonged hazardous situations, by producing a new concept of performing dynamic job rescheduling considering human workforce dynamic disturbances for safety purposes. Managers and workers of the future will realize the full potential of socio-technical digital twins in terms of increases in workers wellbeing and safety.
In order to achieve these long-term results, the researchers involved will investigate the theoretical concept, main architecture, and will develop an experimental prototype at laboratory level. The different technologies will be integrated, tuned and tested and the final expected results of this 34-months project will be as follows:

  1. Development of technical specifications and baselines for human digital twins in manufacturing, comprising a methodological framework for implementing in laboratories and pilot production system a human digital twin. General validation and agreement by all the partners involved.
  2. Digital ergonomics assessment (postural and fatigue evaluation) of manual job execution in manufacturing: development of new proof of concept and testing activities.
  3. Multi-sensor data fusion for workers by laboratory prototypes of Human Digital Twins: integration and testing in the university and company laboratories involved in the project
  4. Experimental data set collection (in laboratories) captured by wearable sensors, biosensors, cognitive and physiological sensors/instruments, development of smart personal protective equipment (PPE) wearable sensors to monitor the job security status and worker health.
  5. Development of new job scheduling models for optimizing human workers involvement and wellbeing in manufacturing and logistics environments, with and without the presence of machines and robots.
  6. Workers’ parameters estimation and wellbeing assessment when passive assistive devices are applied in a manufacturing work station (for instance passive exoskeletons, assistive and adaptive workstations, lifting/handling automatic equipment).
  7. Design of artificial intelligence workflow to predict physiological parameters during working activity
  8. Investigation and report development regarding the expected limits and criticalities in implementing human digital twin at an industrial level in a ordinary working environment and methods/guidelines provision to overcome these limits.
  9. Behavioral Affective Analysis system to map and formalize patterns in workers’ task for identifying best practices according to a human-centered approach and digital twin strategy.
  10. Augmented Reality strategy assisting workers’ tasks for improving safety and efficiency within the workplace
  11. AI-based solutions for analyzing human-computer interactions based on natural language processing (e.g., chat) for estimating and assessing workers’ parameters and wellbeing.
  12. Development of neural-network-based data-driven surrogate models informed by physics computer simulations, for data augmentation of sparse readings coming from wearable devices and prediction of thermal comfort/stress levels.

The final impact of the results before mentioned will be the overall improvement of job quality, health and safety for workers and professional of the Made in Italy system of the future, with particular reference to the manufacturing systems in which humans and machines are collaborating.