21090 research outputs found
Sort by
Beyond the App Store: Reproductive Governance and the Limits of Digital Autonomy
No full textThis paper interrogates how digital contraceptive apps govern reproduction through mechanisms of responsibilization, commodification and consent. While some, like Natural Cycles, are certified as Software as a Medical Device (SaMD), and others, like Clue, emphasize compliance with the European Union's General Data Protection Regulation (GDPR), regulatory protections remain partial and uneven
Assessment of interannual variability of benthic macroinvertebrate assemblages in the Athabasca River
No full textInterannual variability refers to the year-to-year changes in ecological conditions, which can alter niche availability and influence aquatic biota. Benthic macroinvertebrate (BMI) assemblages in riverine ecosystems have been shown to exhibit interannual variability in response to yearly variations in environmental conditions, in particular, large-scale fluctuations involving climate and hydrology. However, knowledge of interannual variation among BMI assemblages in cold region rivers is limited, highlighting the need to better understand year-to-year changes in BMI assemblages and how these changes are associated with environmental variability.
To reduce this knowledge gap, this study quantified interannual changes in BMI assemblages and identified potential associations between hydroclimatic variables and assemblage composition along a cold region river. Benthic macroinvertebrates were collected annually over a range of 6 – 11 years from three reference locations on the mainstem of the Athabasca River, Alberta, Canada. Interannual variability of BMI assemblages was quantified by measuring change in assemblage abundance and taxonomic turnover, along with calculating change values in biological metrics between consecutive years. BMI interannual variability was then related to hydroclimatic variables (climate oscillations, temperature, and discharge) to determine potential environmental drivers.
Results showed that interannual variability of BMI assemblages along the Athabasca River was similar to other studies conducted in cold region streams, and that four taxonomic groups (i.e., Chironomidae, Ephemeroptera, Plecoptera, and Trichoptera) were predominately responsible for yearly changes at all locations. The trajectory of interannual variability and the strength of environmental associations were inconsistent between locations. These results suggest that larger datasets are more likely to provide more reliable quantification of natural variability of BMI assemblages, and other unmeasured factors may influence interannual variability, thus confounding relationships with large-scale environmental drivers. Future studies should consider local-scale variability and finer-resolution environmental predictors to improve our understanding of the drivers of interannual variability
Laurentide Ice Sheet dynamics across multiple glacial-interglacial cycles from Quaternary stratigraphic records in the western Hudson Bay Lowland, central Canada
No full textUnderstanding the terrestrial Quaternary stratigraphic record is necessary to reconstruct regional- to continental-scale paleo-ice sheet fluctuations and compare how these events relate to oxygen isotope proxies derived from marine sediments, sea level change, and contemporary ice sheets. Regions that contain an extensive stratigraphic record beyond the last glacial maximum are key to understanding the long-term behaviour of ice sheets and provide field-based constraints for ice-sheet reconstructions and modelling. The Hudson Bay Lowland (HBL) is one of these key regions situated in central Canada that contains a fragmented stratigraphic record of at least the last four Laurentide Ice Sheet (LIS) glaciations. The main objectives of this thesis are to 1) reconstruct the spatio-temporal evolution of the LIS from the Quaternary stratigraphic record of three relatively understudied regions of the western HBL and 2) determine the timing and climate conditions that persisted during ice-free periods across multiple glacial-interglacial cycles.
To better understand the fragmented stratigraphic record that is dominated by glacial sediments (till), this thesis developed a hybrid lithostratigraphy-allostratigraphy approach that relies on multi-parameter characterization of tills to establish sediment provenance and the ice-flow direction that deposited the sediment. This included detailed study of 70 sections, the collection of 193 stratigraphic ice-flow indicators (154 till fabrics and 39 lodged clasts) and analysis of 393 till samples. Once the till framework was established, the relative age of each nonglacial bed was then determined using the bounding surfaces of designated till units. At least 18 units have been identified this way and correlated across the western HBL. The nonglacial beds were further characterized to assess the timing of sediment deposition using radiocarbon and optical dating methods and paleo-environmental conditions that existed using pollen and foraminifera analysis. Lastly, stratigraphic frameworks presented herein did not rely on geochronology constraints to anchor correlations and the age of identified interglacial beds can be further tested to confirm the age of interglacial beds.
The new stratigraphic framework for the western HBL provides important field-based constraints for LIS reconstructions. This includes evidence that there was asynchronous growth of the two major domes of the LIS during the last two glaciations, with accelerated early growth of the Quebec–Labrador Dome relative to the Keewatin Dome. In each glaciation, the Keewatin Dome becomes more active relatively later in glaciation and persists until deglaciation. During the last glaciation, ice-flowing from the Keewatin Dome likely did not occur until MIS 2 and this S-trending ice-flow transitions into late-glacial SW-trending ice-streams. During the penultimate glaciation (~MIS 6), till deposition by S- to SW-trending ice was extensive and one of the main ice-flow events across the western HBL when the Keewatin Dome was likely situated in northern mainland Nunavut. Furthermore, during deglaciation the Keewatin Dome or an ice divide was likely situated in the western HBL.
There is widespread evidence across the western HBL that sediments belonging to at least three pre-Holocene interglacial periods exists, which provides an important archive to understanding past climatic conditions in central Canada. For two of these interglacials there is evidence of marine inundation which likely occurred during Termination II (~130 ka) and Termination III (~243 ka). In the Churchill River region, the marine limits for both marine incursions are higher compared to the Tyrrell Sea (Holocene) marine limit, providing important field-based constraints for LIS modelling. Importantly, new age estimations from the uppermost intertill nonglacial sediments, combined with consideration of the paleobotanical datasets in the western HBL stratigraphic record, suggest that the region was last deglaciated during MIS 5e. This implies that the HBL, and likely Hudson Bay, remained glaciated during MIS 3.
The presence of weathered bedrock within the western HBL indicates that glacial erosion of bedrock was negligible in places during the Quaternary Period. Relict landscapes across northeastern Manitoba, such as preserved streamlined landform flowsets that are situated outside the margins of late-glacial ice streams, provide evidence of limited erosion following initial glacial advance into the area.
The results of this thesis have provided an updated Quaternary stratigraphic framework for the western HBL, a key region for understanding the long-term evolution of the LIS. The Quaternary stratigraphic record is highly fragmented, which reflects patchy erosion and deposition over multiple glacial-interglacial cycles of the region. The new stratigraphic frameworks developed provide an increased understanding of the growth, evolution and retreat of the LIS during the past two glaciations and insights into pre-Illinoian glaciations, which are essential to improving reconstruction and modelling of the ice sheet throughout the Middle and Late Pleistocene
Compensatory Stepping: Biomechanical Analysis, Predictive Modeling, and Assistance with Lower-Limb Exoskeleton
No full textBalancing is a fundamental human motor function that is continuously performed in daily life. Despite the fact that quiet standing is inherently unstable, humans are able to maintain upright posture through a combination of balance strategies in response to external disturbances. Small perturbations are typically counteracted using ankle and hip strategies, whereas larger and unexpected perturbations often require a reactive stepping response to prevent a fall. Compensatory stepping is a rapid, largely subconscious balance recovery strategy in which a step is taken when the center of mass, or its dynamic equivalent, moves beyond the base of support.
This thesis focuses on the mechanisms underlying compensatory stepping, particularly in response to anterior--posterior impulsive torso perturbations. By investigating the biomechanical and physical principles governing this behavior, the goal is to develop predictive models that can be leveraged for balance assistance in lower-limb exoskeletons, thereby supporting individuals with impaired balance control.
We first conducted human-only compensatory stepping experiments to gain biomechanical insights and identify relationships between stepping parameters (step length and step duration), perturbation characteristics, muscle activation and co-contraction patterns, ground-referenced stability measures, and critical temporal events. Exploratory and regression-based analyses revealed several trends, including relationships between perturbation impulse and stepping probability, non-heel-strike foot contacts, lack of hip balance strategy, and phase-dependent muscle co-contraction strategies. These new findings provide a deeper understanding of the coordination between neuromuscular activity and mechanical stability during reactive stepping.
Building on these insights, a physics-based stepping model was developed to derive a dynamic condition that determines when stepping is required following an impulsive perturbation. A multi-phase, multibody dynamic Compensatory Foot Placement Estimator (CFPE) was then proposed to predict step length and step duration by optimizing multiple cost functions, including stability, joint torque, and restoring effort-related terms. A genetic algorithm was employed for optimization, and the influence of individual cost terms and trajectory generation assumptions was systematically analyzed. Results indicated that the relative importance of cost functions varies with perturbation magnitude and step characteristics.
To overcome the limitations of model-based prediction from a single initial condition and errors due to model simplification, a data-driven convolutional neural network (CNN) was developed to estimate stepping parameters using pre–toe-off exoskeleton kinematic histories as inputs. The instant at which the center-of-mass projection crosses the base of support was identified as a key event containing sufficient predictive information prior to toe-off. The CNN demonstrated robust prediction performance across participants, and recursive feature elimination was used to identify the most influential kinematic inputs.
Finally, two exoskeleton-assisted compensatory stepping controllers—a scaled feedforward controller and a modified velocity flow field (VFF) feedback controller—were implemented on an Indego lower-limb exoskeleton using the predicted stepping parameters. Experimental evaluation using quantitative metrics and participant feedback highlighted the advantages and limitations of each control strategy and demonstrated the feasibility of predictive, parameter-driven balance assistance.
Overall, this work presents a comprehensive framework for understanding, modeling, and assisting compensatory stepping. The findings contribute to improved balance modeling and provide a foundation for the development of intelligent, predictive controllers for rehabilitative and assistive exoskeleton technologies
Towards Green AI: Evaluating Energy Efficiency for Quantized LLM Inference
No full textWhile it is known that quantization of large language models (LLMs) reduces memory usage via lower-bit weights, studies quantifying the resulting impact on energy usage and carbon intensity are scarce. We present the first unified evaluation of weight-only quantization strategies for LLM inference across varying input/output lengths and GPUs, highlighting differences in energy efficiency in addition to accuracy degradation and runtime. We study quantized inference on Llama 2 7B, Phi 3.5 Mini Instruct, OLMo 1B, Qwen 2.5 7B, and Qwen 2.5 14B across GLUE MNLI, MMLU, HumanEval, and GSM8K datasets, evaluating 10 post-training quantization (PTQ) strategies on NVIDIA H100, A6000, A100, and L4 GPUs. We identify a novel trend showing that quantization techniques tend to exhibit peak energy efficiency relative to full-precision baselines when inputs are sufficiently long and outputs are short. Furthermore, we show that quantization strategies become marginally more energy-efficient relative to full-precision models as batch sizes increase, though gains are modest. Notably, fused-kernel implementations such as EETQ int8 and Bitsandbytes int8 offer the highest energy savings, up to 4× compared to FP32 on short text generation, with negligible accuracy loss. Finally, we observe that energy usage closely tracks runtime on our evaluated benchmarks, indicating that, in practice, latency optimization can serve as a reliable proxy for reducing the environmental footprint of LLM services. We conclude with suggested directions for strategically selecting low-carbon quantization strategies based on specific inference requirements
Dissolvable Sugar-Based Untethered Magnetic Millimeter-Scale Robot for Blood Clot Removal
No full textThrombosis, or the formation of blood clots, is a potentially life-threatening condition that results in the complete or partial occlusion of a blood vessel. It remains one of the most prevalent causes of death worldwide. Current treatment approaches involve the intravenous administration of thrombolytic drug, which can increase the risk of uncontrolled bleeding, or catheter-directed treatments, which may have limited access to hard-to-reach areas of the vasculature and can cause catheter-related injuries. Untethered magnetic robots present an alternative approach for thrombosis treatment that addresses the current shortcomings. In this work, a rapidly dissolvable, millimeter-scale, magnetic robot is proposed for the delivery of thrombolytic drug for thrombus removal. The robot is composed of a sucrose-based material with embedded superparamagnetic iron oxide nanoparticles for magnetic control. Although assessment of the robot actuation showed limited forward propulsion from its helical structure, it was able to withstand a maximum flow rate of approximately 72 mL/min, which is comparable to the literature for small-scale magnetic robots. The mean dissolution time of the sucrose structure was 4.65 minutes. The blood compatibility of the robot material was measured through the upregulation of platelet activation markers and found to improve with decreasing material concentration. The drug delivery mechanism consisted of a sealed cavity along the center of the robot to maximize thrombolytic load and avoid drug exposure to high temperatures during fabrication. Release of a placeholder fluorescent protein was found to be gradual across the entire dissolution of the robot. To ensure the thrombolytic agent was not compromised upon loading into the robot, in vitro incubation with human thrombi was performed. The thrombolytic-loaded robot showed similar thrombus mass reduction compared to the direct administration of thrombolytic agent. Finally, the robot functionality was validated using an ex vivo endovascular thrombosis model of the sheep iliac arteries. The robot was clearly visualized under x-ray fluoroscopy due to its embedded magnetic material, enabling its guidance to the ex vivo thrombus via an external rotating permanent magnet mounted on a robotic arm. Successful navigation through a vascular bifurcation was demonstrated and robot mechanical action was shown to accelerate clot mass reduction. The effect of localized thrombolytic delivery on clot mass reduction in the ex vivo model was inconclusive. Overall, the proposed untethered and dissolvable robot would enable the localized delivery of thrombolytic agent to blood clots without the need for retrieval. This can lead to improved patient outcomes by reducing the risks of catheter-related injuries and uncontrolled bleeding resulting from the systemic administration of thrombolytic agents
The Forces Behind the Flux: Methane, Carbon Dioxide, and Nitrous Oxide Dynamics and Their Environmental Drivers in Restored Agricultural Wetlands
No full textWetlands provide vital ecosystem services such as water filtration and flood mitigation but are also significant natural sources of greenhouse gases (GHGs), particularly methane (CH₄). This study examined seasonal and spatial patterns of CH₄, carbon dioxide (CO₂), and nitrous oxide (N₂O) emissions from seven restored agricultural wetlands in the Ontario portion of the Lake Erie Basin, focusing on diffusive and ebullitive flux pathways. Emissions were measured across four seasons, alongside water quality parameters used to identify key environmental drivers. Dissolved oxygen (DO) emerged as a strong driver of GHG fluxes, with lower DO concentrations consistently promoting higher CH₄, CO₂, and N₂O emissions. Duckweed cover also enhanced CH₄ production by creating anoxic conditions. GHG emissions peaked during summer months with heightened biological activity, while winter fluxes, though reduced, remained detectable, emphasizing the contribution of cold-season processes. N₂O emissions remained consistently low throughout the year. Across sites, methane emissions were generally low relative to natural temperate wetlands, except at one nutrient-enriched outlier (Site MA). Spatial variation within wetlands was minimal, suggesting that sampling from a single representative location may be sufficient for long-term monitoring. These findings show that restored agricultural wetlands can act as both carbon sinks and GHG sources depending on local biogeochemical conditions. By identifying major environmental controls on emissions, this study advances understanding of GHG dynamics in restored wetlands, informs efficient monitoring and modelling considerations, and strengthens national inventory and restoration policy development
Beyond the Frame: Culture, Identity, and the Caribbean Sea Diaspora in Canada
No full textThe Caribbean Sea is a complex landscape that holds a storied past and a palimpsest of identities. Indigenous, colonial, and modern traditions shift like tides, and their relationships shape the collective identity of the Caribbean Sea’s inhabitants. Migration from European settlers, the colonial slave trade, and modern refugee crises have created a creolization of cultures and customs. Flows of migration by members of these community have thus created a global archipelago for the Caribbean Sea diaspora who are in constant dialogue and in search for community and representation in their respective exclaves.
These migratory flows have allowed Canada and the Toronto area to become a region which welcomed this diaspora, establishing culture and community. Globalization and the internet have connected us, but in many ways also divided this diaspora from traditional cultural and artistic community activities that must be experienced physically. “Beyond The Frame: Community, Identity, and the Caribbean Sea Diaspora in Canada” attempts to restore the flows of these connections by means of artistic expression and by establishing a central place of community for the Caribbean Sea diaspora in Canada.
Approaching culture and identity from a relational sense, the design for an arts-based centre for immigrants of the Caribbean Sea in Toronto that is rooted in the typologies surrounding the Caribbean Sea takes shape. Cultural institutions are typologies where new relationships are constantly forged between artifacts, people, and space, creating new meaning based on past experiences. This thesis argues that architecture is a tool that helps establish and democratize connections, exploring firsthand accounts from immigrants and their views on identity and culture, leaders from organizations within the Greater Toronto Area, and responsive design concepts in order to produce a representative architectural snapshot of the Caribbean Sea diaspora today
Mobility and the Landscape: Investigating mobility of individuals at Wadi Faynan 100 using minimally invasive strontium isotope analysis
No full textIn comparison to other sites in Jordan, life at the Early Bronze Age (~3600-3000BC) site of Wadi Faynan 100 (WF100) is still largely a mystery. To better understand the use of this site in relation to the EBA transformation of social organization toward urbanism, this study explores strontium isotopic variation using laser ablation-multicollector-inductively coupled plasma-mass spectrometry (LA-MC-ICP-MS) to observe movement to and from the local area. Strontium (Sr) isotope ratios throughout the developmental periods of human enamel were used to indicate locality and patterns of mobility at WF100. Thirty-one samples consisting of a collection of permanent incisors, premolars, and first and third molars, in addition to two deciduous molars were analyzed alongside eleven local faunal samples. Initial results indicate variable patterns of mobility throughout the individual’s childhood development, with some increased instances of consistent locality in later childhood. This supports arguments for a more diversified and regionally specific social organization in EBA Jordan and Wadi Faynan, that may embrace a spectrum of sedentism and transhumance in childhood. These results highlight both the potential for, and challenges of conducting further LA-MC-ICP-MS analysis of Sr in the broader Jordan landscape and provide novel insights into EBA mobility using sequential dental sampling
Bayesian Inference for Partial Differential Equations via Neural Network Surrogates
No full textPartial differential equations (PDEs) provide the fundamental framework for describing physical systems; yet, in many practical applications, these equations contain unknown parameters that must be inferred from experimental observations. Solving such inverse problems using traditional mesh-based numerical methods is often computationally intensive; furthermore, because these solvers cannot be easily differentiated with respect to model parameters, they create significant bottlenecks for gradient-based inference. To address these challenges, we train parameterized Physics-Informed Neural Networks (PINNs) for two distinct systems: the Allen–Cahn and Cahn–Hilliard (AC–CH) phase field equations and diffusion models for cyclic voltammetry (CV). These surrogates demonstrate strong generalizability across continuous parameter spaces and serve as differentiable components for gradient-based Bayesian parameter estimation via the No-U-Turn Sampler (NUTS). This work verifies the feasibility of a unified PINN-surrogate-Bayesian workflow for parameter estimation, offering a promising complement to existing methods for solving inverse problems with uncertainty quantification