Giulia Bassani

Dr. Giulia Bassani

orcid

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
11 July 2020 in Sensors
Verified
Journal Article
Open access
PDF file
PGPaolo Giannini
Paolo Giannini
GBGiulia Bassani
Giulia Bassani
Carlo Alberto AvizzanoCarlo Alberto Avizzano
Carlo Alberto Avizzano
AFAlessandro Filippeschi
Alessandro Filippeschi
The assessment of risks due to biomechanical overload in manual material handling is nowadays mainly based on observational methods in which an expert rater visually inspects videos of the working activity. Currently available sensing wearable technologies for motion and muscular activity capture enables to advance the risk assessment by providing reliable, repeatable, and objective measures. However, existing solutions do not address either a full body assessment or the inclusion of measures for the evaluation of the effort. This article proposes a novel system for the assessment of biomechanical overload, capable of covering all areas of ISO 11228, that uses a sensor network composed of inertial measurement units (IMU) and electromyography (EMG) sensors. The proposed method is capable of gathering and processing data from three IMU-based motion capture systems and two EMG capture devices. Data are processed to provide both segmentation of the activity and ergonomic risk score according to the methods reported in the ISO 11228 and the TR 12295. The system has been tested on a challenging outdoor scenario such as lift-on/lift-off of containers on a cargo ship. A comparison of the traditional evaluation method and the proposed one shows the consistency of the proposed system, its time effectiveness, and its potential for deeper analyses that include intra-subject and inter-subjects variability as well as a quantitative biomechanical analysis.

15 September 2021 in IEEE Sensors Journal
Verified
Journal Article
Closed access
GBGiulia Bassani
Giulia Bassani
AFAlessandro Filippeschi
Alessandro Filippeschi
CACarlo Alberto Avizzano
Carlo Alberto Avizzano
Manual Material Handling (MMH) activities represent a large portion of the workers’ tasks in the tertiary sector. The ability to monitor, model, and predict human behaviours are crucial to both the design of productive human-robot collaboration and an efficient physical exposure assessment system that can prevent Work-related Musculoskeletal Disorders (WMSDs), with the ultimate goal of improving workers’ quality of life. The combined use of wearable sensors and machine learning (ML) techniques can fulfil these purposes. Inertial Measurement Units (IMUs) and surface Electromyography (sEMG) allow collecting kinematic data and muscular activity information that can be used for biomechanical analyses, ergonomic risk assessment, and as input of ML algorithms aimed at joint torque/load estimation, and Human Activity Recognition (HAR). The latter needs a large amount of annotated training samples, and the use of publicly available datasets is the way forward. Nowadays, the majority of them concern Activities of Daily Life (ADLs) and, including only kinematic data, have limited applications. This paper presents a fully labelled dataset of working activities that include full-body kinematics from 17 IMUs and upper limbs sEMG data from 16 channels. Fourteen subjects participated in the experiment performed in laboratory settings for overall 18.6 hours of recordings. The activities are divided into two sets. The first includes lifting, lowering, and carrying objects, MMH activities suitable for ergonomic risk assessment, and HAR. The second includes isokinetic arm movements, mainly targeting load and joint torque estimation.

19 November 2024 in Robotics
Verified
Journal Article
Open access
PDF file
Riccardo BezziniRiccardo Bezzini
Riccardo Bezzini
GBGiulia Bassani
Giulia Bassani
Carlo Alberto AvizzanoCarlo Alberto Avizzano
Carlo Alberto Avizzano
AFAlessandro Filippeschi
Alessandro Filippeschi
The recent advancements in wearable exoskeletons have highlighted their effectiveness in assisting humans for both rehabilitation and augmentation purposes. These devices interact with the user; therefore, their actuators and power transmission mechanisms are crucial for enhancing physical human–robot interaction (pHRI). The advanced progression of 3D printing technology as a valuable method for creating lightweight and efficient gearboxes enables the exploration of multiple reducer designs. However, to the authors’ knowledge, only sporadic implementations with relatively low reduction ratios have been reported, and the respective experimental validations usually vary, preventing a comprehensive evaluation of different design and implementation choices. In this paper, we design, develop, and examine experimentally multiple 3D-printed gearboxes conceived for wearable assistive devices. Two relevant transmission ratios (1:30 and 1:80) and multiple designs, which include single- and double-stage compact cam cycloidal drives, compound planetary gearboxes, and cycloidal and planetary architectures, are compared to assess the worth of 3D-printed reducers in human–robot interaction applications. The resulting prototypes were examined by evaluating their weight, cost, backdrivability, friction, regularity of the reduction ratio, gear play, and stiffness. The results show that the developed gearboxes represent valuable alternatives for actuating wearable exoskeletons in multiple applications.

19 August 2024 in Sensors
Verified
Journal Article
Open access
PDF file
CCCristian Camardella
Cristian Camardella
VLVittorio Lippi
Vittorio Lippi
FPFrancesco Porcini
Francesco Porcini
GBGiulia Bassani
Giulia Bassani
LLLucia Lencioni
Lucia Lencioni
AFAlessandro Filippeschi
Alessandro Filippeschi
Using lower limb exoskeletons provides potential advantages in terms of productivity and safety associated with reduced stress. However, complex issues in human–robot interactions are still open, such as the physiological effects of exoskeletons and the impact on the user’s subjective experience. In this work, an innovative exoskeleton, the Wearable Walker, is assessed using the EXPERIENCE benchmarking protocol from the EUROBENCH project. The Wearable Walker is a lower-limb exoskeleton that enhances human abilities, such as carrying loads. The device uses a unique control approach called Blend Control that provides smooth assistance torques. It operates two models simultaneously, one in the case in which the left foot is grounded and another for the grounded right foot. These models generate assistive torques combined to provide continuous and smooth overall assistance, preventing any abrupt changes in torque due to model switching. The EXPERIENCE protocol consists of walking on flat ground while gathering physiological signals, such as heart rate, its variability, respiration rate, and galvanic skin response, and completing a questionnaire. The test was performed with five healthy subjects. The scope of the present study is twofold: to evaluate the specific exoskeleton and its current control system to gain insight into possible improvements and to present a case study for a formal and replicable benchmarking of wearable robots.

2 November 2025 in Sensors
Verified
Journal Article
Open access
PDF file
GBGiulia Bassani
Giulia Bassani
Carlo Alberto AvizzanoCarlo Alberto Avizzano
Carlo Alberto Avizzano
AFAlessandro Filippeschi
Alessandro Filippeschi
Human Activity Recognition (HAR) is widely used for healthcare, but few works focus on Manual Material Handling (MMH) activities, despite their diffusion and impact on the workers’ health. We propose four Deep Learning algorithms for HAR in MMH: Bidirectional Long Short-Term Memory (BiLSTM), Sparse Denoising Autoencoder (Sp-DAE), Recurrent Sp-DAE, and Recurrent Convolutional Neural Network (RCNN). We explored different hyperparameter combinations to maximize the classification performance (F1-score,) using wearable sensors’ data gathered from 14 subjects. We investigated the best three-parameter combinations for each network using the full dataset to select the two best-performing networks, which were then compared using 14 datasets with increasing subject numerosity, 70–30% split, and Leave-One-Subject-Out (LOSO) validation, to evaluate whether they may perform better with a larger dataset. The benchmarking network DeepConvLSTM was tested on the full dataset. BiLSTM performs best in classification and complexity (95.7% 70–30% split; 90.3% LOSO). RCNN performed similarly (95.9%; 89.2%) with a positive trend with subject numerosity. DeepConvLSTM achieves similar classification performance (95.2%; 90.3%) but requires ×57.1 and ×31.3 more Multiply and ACcumulate (MAC) and ×100.8 and ×28.3 more Multiplication and Addition (MA) operations, which measure the complexity of the network’s inference process, than BiLSTM and RCNN, respectively. The BILSTM and RCNN perform close to DeepConvLSTM while being computationally lighter, fostering their use in embedded systems. Such lighter algorithms can be readily used in the automatic ergonomic and biomechanical risk assessment systems, enabling personalization of risk assessment and easing the adoption of safety measures in industrial practices involving MMH.

27 November 2025 in IEEE Access
Verified
Journal Article
Open access
Riccardo BezziniRiccardo Bezzini
Riccardo Bezzini
GBGiulia Bassani
Giulia Bassani
CACarlo Alberto Avizzano
Carlo Alberto Avizzano
AFAlessandro Filippeschi
Alessandro Filippeschi
The recent advancements in assistive exoskeletons have highlighted their effectiveness in reducing the users’ metabolic consumption, providing functional compensation, and preventing biomechanical overloads. The exoskeletons’ transmission mechanisms are essential for guaranteeing their effectiveness, safety, and comfort. Therefore, lightweight, efficient, backdriveable, and minimally bulky reduction gearboxes with low mechanical output impedance have become a widespread choice in modular assistive exoskeletons. In previous works, the value of the cycloid-planetary series transmission has been assessed. However, the previous design turned out to be too bulky in the axial direction. In this paper, we propose a novel CYcloidal-PLanetary cOncentric series combination (CYPLO) that exploits a compact-cam cycloidal architecture nested inside the sun of a planetary stage. This design enables valuable performance regarding key gearbox features such as regularity, friction, and gear play, while maintaining an extremely advantageous compactness in the axial direction and significant backdriveability even for high transmission ratios. A reducer implementing this novel solution has been designed, developed through 3D printing, and examined experimentally to evaluate its performance with respect to the previous designs and state-of-the-art gearboxes employed in wearable devices’ actuators. The results show that this transmission architecture represents a valid solution for lightweight assistive exoskeletons, allowing for high reduction ratios while maintaining compactness and backdriveability.