University of Quebec

To develop an effective quasi-passive (QP) exoskeleton, maximizing its range of variable assistive torques while minimizing the energy required for this variation is crucial. However, achieving this goal has proven challenging so far owing to the common trade-off between the range of variable assistive torques and the energy required for torque variation. Additionally, the shoulder’s standby posture complicates the search for viable solutions. To tackle this issue, we derived design principles based on elastic potential energy field. Utilizing these principles, we developed a QP shoulder exoskeleton called adjustable shoulder exoskeleton (AD Exo), which successfully achieved a wide range of variable assistive torques with significantly reduced required energy for variation. Benchtop tests demonstrated a remarkable enhancement in variability, with a range of assistance spanning 6.37 Nm achieved with a variation energy of 0.9 J. In human trials, AD Exo significantly minimized the average percentage of maximum voluntary contraction in shoulder muscles. Compared to the condition without the exoskeleton (NE), the average muscle activation was reduced by 25% at the adjusted assistance (ADJ), 7.5% at the low assistance (LOW), and 6.7% at the high assistance (HIGH), respectively. Considering repetitive and long-term tasks, this reduction in muscle activity can accumulate, making AD Exo highly effective for alleviating shoulder muscle burden and fatigue. Furthermore, kinematic motions of wearers and actual assistive torque delivered to wearers were examined to analyze the underlying the assistive effect of the exoskeleton.

High gain antennas play an essential role in future wireless applications by ensuring sufficient compensation for propagation losses in long-range communication. This work presents a novel design of metasurface-inspired high gain miniaturized antenna for 28 GHz millimeter wave applications. Firstly, to improve the radiation gain of the suggested antenna, only two metallic shorting pins are inserted between radiating patch and ground of antenna.The total size of the proposed shorting pin based microstrip patch antenna is 18$\times$14 mm². This approach differs from traditional methods of increasing the radiating area or using bulky techniques. For further gain enhancement, the antenna incorporates a single layer of metasurface made up of a compact and uniquely designed reflecting metamaterial unit cell array. The distinctiveness of these unit cell design is centered on a simple 2$\times$2 array created by combining circular and hexagonal split rings. This innovative configuration improves antenna gain while maintaining a compact form factor.This metamaterial array is positioned at approximately 0.5 $\lambda _{0}$ (where $\lambda _{0}$ is the wavelength in millimeters) above the antenna, thus creating a Fabry-Perot cavity, and substantially improves the gain exceeding 13 dBi.The metasurface, which is made up of specially designed unit cells, modifies the electromagnetic field distribution and surface waves to produce better radiation properties. Furthermore, the final optimized bandwidth achieved is 26.62–29.79 GHz with peak radiation efficiency up to 94%, respectively. The findings provide satisfactory agreement when comparing simulation results with experimental data. The presented antenna system is developed to meet stringent gain requirements for 5G antennas operating in the millimeter wave frequency band.

Optical and mechanical properties of europium (Eu)-doped silicon nitride (SixNy) films were investigated as a function of the sputtering power applied to the Eu metal target and argon flow into the deposition chamber. Films were fabricated by an electron cyclotron resonance plasma-enhanced chemical vapor deposition (ECR-PECVD) system combined with magnetron sputtering for in situ rare-Earth doping. Results show that Eu-doped SixNy films fabricated with higher sputtering power exhibit intense red emission when annealed above 1000 °C, making the luminescence visible under daylight conditions. Variable-angle spectroscopic ellipsometry analysis shows that the refractive index and film thickness are strongly dependent on the sputtering power; however, argon (Ar) flow has minimal influence. High-resolution X-ray diffraction reveals that the crystalline phases in the films play a crucial role in efficient Eu emission. Additionally, the films show increased hardness, up to 18 GPa, and an elastic modulus of 160 GPa, ensuring their durability and performance as cladding layers in photonic devices. These mechanical properties are essential for maintaining structural integrity and preventing defects, which are critical for the reliability of optoelectronic devices. The combination of strong emissions and good mechanical properties make Eu-doped SixNy films suitable for optoelectronic and solar cell applications, where both efficient light emission and material stability are essential.

Federated learning (FL) is a distributed learning framework designed for large-scale applications. The core advantage of FL is that each participant is not required to share the local data with a central server. This inherent privacy-preserving capability is well suited to the increasingly popular large-scale generative AI models. However, some studies have shown that FL is susceptible to reconstruction attacks, in which an attacker leverages the acquired gradients or model parameters to reconstruct a victim's data. In this article, we propose a novel reconstruction attack scheme based on generative adversarial networks (GANs) in an asynchronous FL scenario that can reconstruct a victim's dataset without an auxiliary dataset. This adversarial scheme demonstrates a significant ability to reconstruct a dataset of victims accurately, thereby posing a substantial threat to user privacy, particularly in the context of large-scale models. Furthermore, we explore various defense mechanisms based on the characteristics of asynchronous FL and ultimately establish a viable defense scheme based on homomorphic encryption and an intermediate server. The proposed defense framework successfully and flawlessly defends against reconstruction attacks from the server side in an asynchronous setting without degradation of the model performance.

In this paper, we present a multimodal dataset obtained from a honey bee colony in Montréal, Quebec, Canada, spanning the years of 2021 to 2022. This apiary comprised 10 beehives, with microphones recording more than 3000 hours of high quality raw audio, and also sensors capturing temperature, and humidity. Periodic hive inspections involved monitoring colony honey bee population changes, assessing queen-related conditions, and documenting overall hive health. Additionally, health metrics, such as Varroa mite infestation rates and winter mortality assessments were recorded, offering valuable insights into factors affecting hive health status and resilience. In this study, we first outline the data collection process, sensor data description, and dataset structure. Furthermore, we demonstrate a practical application of this dataset by extracting various features from the raw audio to predict colony population using the number of frames of bees as a proxy.

Objective: The story recall subtest of the Batterie d'Efficience Mnésique (BEM-144) is a verbal episodic memory test that assesses immediate and episodic memory. Variables such as age, sex, and education level can impact performance on this type of memory test, as can cultural differences. Therefore, the purpose of this study was to establish normative data for the story recall subtest of the BEM-144 in the elderly French-Quebec population. Method: The normative sample consisted of 260 healthy individuals aged 50-90 years, all from the province of Quebec, Canada. Analyses were performed to estimate the association between age, sex, and education level on one hand, and immediate and delayed recall performance, on the other hand. Results: The results show that all sociodemographic variables are significantly associated with story recall performance. Normative data are proposed in the form of regression equations. Conclusions: Overall, these norms will be beneficial for the evaluation and detection of episodic memory impairment in middle-aged and older adults.

Verbal fluency tests are used to assess executive functions and language. The verb fluency test has proven successful in distinguishing healthy individuals (HCs) from participants with pathological conditions. However, few normative and psychometric studies have been published for the verb fluency test. The aim of Study 1 was to provide normative data in the adult population of French Québec for the verb fluency test. The aim of Study 2 was to determine its discriminant validity and test-retest reliability. The normative sample consisted of 424 HCs aged 50-92 years. Multiple linear regressions were used to generate equations for calculating Z-scores. To assess discriminant validity, the performance of 46 HCs was compared with that of 46 participants with mild cognitive impairment (MCI). To determine test-retest reliability, the test was administered twice, 3 months apart, to a group of 25 HCs. Age, sex, and education level were significantly related to performance on the test. The test distinguished the performance of HCs from that of participants with MCI. Test-retest analysis showed that scores had good stability over time. Norms and psychometric data for the verb fluency test will help clinicians and researchers better identify executive and language impairments associated with pathological conditions.

Objective: The 12-item Buschke memory test is used to assess verbal episodic memory in adults and older adults. However, there is no normative data for this test adjusted to the older Quebec-French population. The aim of the study was to produce normative data for the 12-item Buschke for the Quebec-French population aged 50 and older. Method: The normative sample consisted of 172 healthy French-speaking participants aged 50-89 years, from the Province of Quebec (Canada). The influence of age, years of formal education, and sex on five 12-item Buschke scores were analyzed. Based on the distribution of scores, normative data were developed as Z-scores equation, regression equation, and percentiles. Results: Age, years of formal education, and sex were all associated with performance. Equations to calculate Z-scores were provided for the free recall trial 1 and the free recall trials 1-3. Stratified percentiles were provided for the delayed free recall and total recall 1-3. Conclusions: The normative data for the 12-item Buschke improve the accuracy of clinicians to detect verbal episodic memory impairments in Quebec's aging population.

Due to growing environmental concerns, climate change, and the scarcity of natural resources, there have been recent efforts to produce chemicals from lignocellulose. The objective of this study was to valorize cocoa pods husks into lactic acid. Delignification of the raw substrate was carried out with potassium hydroxide, with a biomass/solvent ratio of 9.14% and a temperature of 128 °C, giving a maximum delignification of 93.87%. The acid hydrolysate obtained in 26.95 min of autoclave, with a biomass/solvent ratio of 9.53% (m/v) and a concentration of 4.92% (m/v) of H2SO4 after detoxification and neutralization was fermented into lactic acid by Lactobacillus fermentum ATCC 9338 in a LAMBDA MINIFOR biofermentor, with control of pH, O2, temperature and stirring speed. A content of 18.77 g L− 1 of reducing sugars, a productivity of 1.25 g/(L.h), a titre of 26.61 ± 0.61 g L− 1 and a yield of 0.52 g g− 1 of lactic acid in the logarithmic phase after 72 h, from 45.55 g L− 1 of glucose and 12.21 g L− 1 of xylose. Tolerance to inhibitors (phenolic and furfural compounds) has been observed.

High-energy ultraviolet (UVC) irradiation of metal oxides (MOs, e.g., TiO2) results in photoinduced surface oxygen vacancies (PI-SOVs), which can change the charge carrier (e.g., electrons and holes) migration dynamics. Although PI-SOVs alter the electronic and chemical properties of MOs, there is no consensus on the penetration depth of the UVC irradiation, which induces PI-SOVs and is an important variable for the design and operation of MO-based systems. Here, we performed optical transmission and time-resolved atomic force microscopy measurements on back-illuminated TiO2 samples. Our experiments show that the effect of UVC irradiation on MOs can be observed hundreds of micrometers across the bulk, i.e., orders of magnitude larger than previously postulated values. We believe that our findings would be important both for the fundamental understanding of UVC irradiation/penetration and for device design/fabrication processes.

Wood density is a critical control on tree biomass, so poor understanding of its spatial variation can lead to large and systematic errors in forest biomass estimates and carbon maps. The need to understand how and why wood density varies is especially critical in tropical America where forests have exceptional species diversity and spatial turnover in composition. As tree identity and forest composition are challenging to estimate remotely, ground surveys are essential to know the wood density of trees, whether measured directly or inferred from their identity. Here, we assemble an extensive dataset of variation in wood density across the most forested and tree-diverse continent, examine how it relates to spatial and environmental variables, and use these relationships to predict spatial variation in wood density over tropical and sub-tropical South America. Our analysis refines previously identified east-west Amazon gradients in wood density, improves them by revealing fine-scale variation, and extends predictions into Andean, dry, and Atlantic forests. The results halve biomass prediction errors compared to a naïve scenario with no knowledge of spatial variation in wood density. Our findings will help improve remote sensing-based estimates of aboveground biomass carbon stocks across tropical South America.

Dismounted situational awareness (DSA) is a critical component of military operations. It is enhanced by tactical augmented reality (TAR) systems that overlay digital information onto soldiers' physical environments. Traditional TAR systems rely predominantly on data from soldier-mounted cameras, which can limit their effectiveness and increase the risk of soldiers being exposed to unseen threats. To address these challenges, we propose a new TAR framework called Tactical Augmented Reality on the Move (TAROTM). TAROTM utilizes advanced military robots, such as quadruped unmanned ground vehicles (QUGVs), that are organized into specialized collaborative teams to support sensing, data processing, storage, and analytics. Given the significant volume of data, amount of traffic, and delay constraints associated with TAROTM, we explore quantum machine learning (QML)'s potential to enable real-time data processing, analytics, and distribution. As a case study, we employ QML to optimize sensor-to-shooter data routing in TAROTM. Additionally, we discuss the challenges and opportunities associated with integrating QML in the TAROTM system.

Mutations on the Ras-family of small GTPases are among the most common molecular oncogenic drivers, with the HRas isoform being primarily associated with head-and-neck and genito-urinary cancers. Although once considered “undruggable,” recent efforts have identified a structurally conserved surface pocket in the Ras family, designated the SI/II pocket, situated near the binding site of the guanidine exchange factor (GEF) SOS1. The SI/II pocket may represent a potential target site for a pan-Ras drug. A crystal structure representing the native state of GDP-bound HRasG12V was generated to characterize the topology of the SI/II pocket. This native-state structure was employed, together with the published structure of GppNHp-bound HRasG12V in state 1 (PDB ID: 4EFM), as a base for further molecular dynamics simulations exploring the conformational dynamics of the SI/II pocket via four generated synthetic HRas model structures. Our results show that the SI/II pocket is natively inaccessible in GDP-bound HRas yet becomes accessible in state 1 GppNHp-bound HRas systems, an effect that seems to be more evident in the mutated enzyme. This points to the GTP-bound state as a most promising target for Ras inhibitors directed at the SI/II pocket. Occlusion of the SI/II pocket is dictated by the spatial position of the α2 α helix in relation to the protein core, with α2 residue Y71 acting as a “tyrosine toggle” capable of restricting the pocket access.

Vibration energy harvester has always been a subject of interest of research, highlighting the importance of understanding its behavior and explore their potential applications. Piezoelectric and magnetostrictive harvesters are particularly interesting due to their high-power density extraction capabilities. This paper presents an analytical method for obtaining and comparing important output parameters of a piezoelectric-magnetostrictive energy harvester based on the strain energy approach. This method offers a straightforward and simple enough to compare the impact of different geometric and physical parameters on key outcomes such as power, voltage, and current. The harvester beam consists of a metal base surrounded by piezoelectric and magnetostrictive layers subjected to base excitation. Initially, direct relationships were established to calculate the load and unload natural frequencies. Subsequently, the interface circuit was investigated in three configurations: "separate", "parallel", and "series". The effect of layer thicknesses, resistance values, and also pick-up coil turns was examined, and their optimum values were calculated. Due to the capacitive behavior of the piezoelectric layer and the inductive behavior of the magnetostrictive layer, when combined in an equivalent circuit, leads to specific phenomena such as resonance, which have been thoroughly investigated to maximize energy extraction from the hybrid harvester.