7903 research outputs found
Sort by
Microfluidic Intracranial Pressure Monitoring Sensor
Get PDFThis thesis presents the design of a novel microfluidic sensor for continuous monitoring of intracranial pressure (ICP), intended as a less invasive alternative to current monitoring techniques. The sensor targets patients with critical and chronic neurological conditions, aiming to provide real-time data on ICP fluctuations. The microfluidic intracranial pressure sensor is designed for implantation beneath the skull, on the dura mater, and integrates a PDMS-based microfluidic strain sensor with wireless data transmission. Initial testing focuses on optimizing sensor design and integration with a microcontroller. Preliminary benchtop testing demonstrates the sensor\u27s capability to detect ICP with a minimum sensitivity of 5 mbar. Wireless data transmission was integrated by using a microcontroller to turn resistance measurements into readable pressure values using a voltage divider circuit and calibration equation. This cost effective alternative to ICP detection gives real-time data which may improve clinical decisionmaking, enable early detection of ICP changes, and reduce the need for invasive procedures and MRI-associated complications
I-280/Wolfe Road Interchange Redesign
Get PDFWith the increasing population, there is an increasing need for traffic efficiency and road safety. Roadways that were once considered adequate may be unable to accommodate the growing number of vehicles on the road. In this paper, we analyze the I-280/Wolfe Road Interchange to determine its safety and efficiency inadequacies and propose an improved interchange design to fix these issues. Our analysis reveals that there is room for improvement in the safety of drivers, bikers, and pedestrians, as well as opportunities to enhance the efficiency of sections of the road that require it. We identified the lack of protective barriers on Wolfe Road, as well as a high traffic demand in the northern intersection in the southbound direction, to be major issues to address. To address these issues, we designed a new interchange that adds an additional lane to Wolfe Road in the southbound direction and adds barriers between the vehicle lane and the bike lane. With these improvements, the roadway will be safer and more efficient, keeping vulnerable users protected and accounting for future roadway demand
Efficient Inference Of Performance Models in OpenMP Applications
No full textChoosing the best OpenMP parameters such as thread count, scheduling type, and chunk size is essential for optimizing parallel program performance. One of the promising approaches is to infer a (pre-trained) performance model at runtime to determine the parameters to run OpenMP parallel regions. Such a performance prediction model can require programs’ code information such as intermediate representation (IR) and other at-runtime information (e.g., input sizes) to make a performance prediction. In such a scenario, extracting or querying the IR information at runtime can create a significant runtime overhead. This thesis proposes a compiler-asssited tuning framework that shifts IR extraction and instrumentation to compile time using custom LLVM passes, embedding static code features and enabling lightweight runtime inference just before parallel region execution. The approach reduces runtime overhead, and supports selective tuning based on input size and problem complexity. Experimental evaluation on Polybench and NPB benchmarks shows that compiler-assisted inference substantially reduces runtime overhead compared to runtime-only methods, making autotuning practical at scale. Optimizations including a Cython-based inference backend, batched inference calls, and static filtering further enhance efficiency
A Retrieval-Augmented Generation Framework for LEED Compliance in Sustainable Building Design and Construction
No full textAs sustainable construction continues to gain prominence, the industry’s increased emphasis on environmental responsibility and resource efficiency has spurred growing interest in advanced technologies and frameworks. Designing sustainable buildings involves the integration of diverse and interdependent strategies spanning building systems, lifecycle analyses, material selection, and stakeholder behavior. Retrieval-augmented generation (RAG) is a powerful artificial intelligence framework combining the strengths of information retrieval and generative models to enhance contextual understanding and knowledge accessibility. Despite its wide application in various domains, the potential of RAG remains largely unexplored in civil engineering contexts. To bridge this gap, this thesis introduces a tailored RAG framework designed specifically to address sustainability-related queries, significantly improving the efficiency of accessing specialized knowledge for industry practitioners. Leveraging a domain-specific corpus derived from reference materials within the Leadership in Energy and Environmental Design (LEED) rating system, the proposed framework enhances retrieval effectiveness and answer interpretability through the integration of structured knowledge graphs and refined prompt engineering. Iterative optimization of the retrieval and prompting methods further strengthens the system’s ability to handle complex, calculation-intensive queries that are prevalent in sustainable design scenarios. Experimental results demonstrate substantial improvements in both accuracy and practical usability, supporting more informed and reliable decision-making processes in sustainable building projects. Ultimately, this research highlights the transformative potential of AI-powered approaches for streamlining LEED compliance and advancing the field of sustainable construction
