Smart Mattress Pressure Mapping: Machine Learning-Based Sleep Posture, Biomechanical Characterization and (Dis)Comfort Analysis

Institution: University of Salerno; Laboratory Human Centred Design and Vehicle Design by Simulation

Website: https://www.unisa.it/

Country: Italy

Duration: 3–6 months (recommended minimum: 4 months)

Contact information at the placement: Rosaria Califano, rocalifano@unisa.it

Contact information at the higher education institution: bbarone@unisa.it

Description:

This traineeship focuses on the development of an intelligent pressure-based monitoring system for sleep posture, biomechanical characterization and comfort analysis using instrumented mattresses.
The project is entirely centred on mattress–human interaction during sleep and aims to objectively characterize different sleep postures through high-resolution pressure acquisitions, machine learning optimization techniques, integrating biomechanical indicators with subjective (dis)comfort evaluations.

The research adopts a data-driven analytical framework combining:

• High-resolution pressure distribution data acquired through instrumented mattresses
• Biomechanical segmentation of body regions (head, torso, pelvis, upper limbs, lower limbs)
• Quantitative pressure metrics extraction (peak pressure, contact area, center of pressure, symmetry indices)
• Machine Learning algorithms for automatic posture classification
• Correlation analysis between objective pressure metrics and subjective comfort/discomfort ratings

Trainees will:

• Acquire and preprocess large datasets of pressure maps recorded in different sleep postures (supine, lateral, prone, semi-elevated, etc.)
• Develop algorithms to segment pressure maps into anatomical regions
• Extract biomechanical features describing pressure distribution patterns
• Collect and integrate subjective comfort/discomfort scores using structured evaluation scales
• Perform statistical comparison between postures and body regions
• Design and train Machine Learning models for posture recognition
• Evaluate model performance using training/validation splitting and cross-validation techniques
• Investigate the relationship between pressure distribution features and perceived comfort
• Provide biomechanical interpretation of results in the context of sleep ergonomics and pressure injury prevention
• The traineeship aims to bridge objective biomechanical measurements and intelligent classification systems, contributing to the development of next-generation smart mattresses for healthcare, sleep monitoring, and pressure ulcer prevention.

By the end of the traineeship, participants will deliver:

• A structured and documented pressure map dataset
• A posture classification and characterization model
• A biomechanical pressure distribution analysis report
• A statistical validation of model performance
• A correlation analysis between pressure metrics and subjective (dis)comfort
• Engineering recommendations for smart mattress design optimization
• A final technical presentation of results

Requirements:

• Undergraduate or postgraduate students in Mechanical Engineering, Biomedical Engineering, Industrial Engineering, Data Science, or related fields
• Strong interest in biomechanics, ergonomics, and human–machine interaction
  Ability to analyze multidimensional datasets
• Basic knowledge of statistics and Machine Learning techniques
• Familiarity with MATLAB, Python, or similar analytical tools
• Interest in healthcare applications and smart sensing technologies
• English communication skills at B1 level or higher
• Analytical mindset and structured problem-solving approach

Language of communication: English

(Internship/Position ID: NSMT_IT0021)