Past Projects

Our project focuses on developing a web interface to dynamically retrieve and display media assets from MediaValet, a Digital Asset Management (DAM) system. This interactive platform allows users to explore artwork across various albums, offering features like categorized browsing, a randomized gallery, and keyword-based search functionality. Users can search by image attributes or keywords within featured albums or specific collections, enhancing their content discovery experience. Designed for DocArts, the interface emphasizes an engaging and intuitive user experience, showcasing their media in an interactive format. The front-end leverages Next.js to deliver responsive design and visually appealing layouts, while the backend, built with JavaScript, manages seamless content retrieval and API integrations. This platform not only highlights the artistic work but also provides an efficient and enjoyable way to navigate DocArts' extensive media collection.

WillCall is a voice-activated nurse call device designed to provide patients with an accessible and dignified method of requesting assistance. The purpose of the project is to enhance independence and improve response times in healthcare settings by offering a simple, intuitive alternative to traditional nurse call buttons. WillCall utilizes speech recognition technology to detect key phrases without the use of an internet connection, ensuring reliable activation of the system when help is needed completely offline. The device ensures reliable phrase detection, integrates seamlessly with existing nurse call systems, and removes barriers for patients with limited mobility. It enhances patient safety and improves communication in care environments.

During a surgical operation, surgeons often refer to images and prior scans to guide them to create a more effective and efficient surgery. However, one of the major drawbacks of going through these scans currently is having to break sterility to manually manipulate or change scans. Once sterility is broken inside the operating room, it becomes a hassle to sterilize again and can take minutes to get back to the surgery wasting time. To solve this problem, this portable gesture recognition device tracks the hands of the surgeon within a field away from the patient to manipulate the scans and 3D-reconstructions within the 3D slicer program. The program works with 90% accuracy and within 0.2 seconds of the gesture used, creating a seamless and instant result.

Endotracheal (ET) intubation in infants born at 24-25 weeks’ gestation is particularly challenging with current laryngoscope models, yielding first-attempt success rates of 40-60%, significantly lower than the 90%+ success rate seen in adults. A significant factor is the difficulty in placing the ET tube in under 30 seconds, after which infants face the risk of adverse events due to reduced oxygen levels. An improper view of the trachea can also necessitate second or third intubation attempts, leading to trauma to surrounding tissue. The novel neonatal video laryngoscope provides continuous oxygen delivery, integrated suction for secretion removal, and a Computer Vision- enabled targeting system for enhanced tracheal visualization. These features will improve intubation success rates and enhance outcomes for premature infants. The final device includes an optimized blade design for oxygen and suction flow and a Graphical User Interface that assists physicians in quickly and accurately placing the ET tube.

Our team developed a system to improve lung transplant outcomes by automating the process of matching donor and recipient lung sizes using CT image analysis. The system utilizes pre-trained convolutional neural networks to segment lung regions from CT scans and extract volumetric data. This data is then processed using conditional logic algorithms in MATLAB, which account for lung length and donor-recipient compatibility thresholds. This method addresses the challenges associated with size mismatches, which can lead to post-transplant complications. Working closely with the client, we integrated clinical feedback to refine the system and ensure alignment with real- world transplant needs. The final solution enhances the organ allocation process by providing accurate size predictions, improving equity and efficiency in lung distribution.

The Geoscience Department at The University of Texas at Dallas (UTD) is looking to design a sediment catcher that solves the current issue of losing sediment during extraction. The team has designed a core catcher that stops fine water-laden sediment from escaping the bottom of the inserted tube up to five meters in length. Sediment Solutions has designed L.O.C.C. (Load-Operated Core Catcher), a system that utilizes replaceable sheet metal gates that close at the beginning of the extraction phase due to the load of sand that collapses the gate. This is paired with a custom coupling system that connects pipe lengths and two fixtures to support in-field testing. The coupling and housing have been tested in rugged conditions to ensure durability. Similarly, the gate has been tested with various sand particle sizes to verify that it can capture water-laden sand.

The new MagneMotion Automated Material Handling System at Texas Instruments consists of a fleet of magnetic carts propelled along a track via alternating magnetic forces. To keep the tracks clean, Texas Instruments has tasked UTDesign Team 2044 (JAMBER Kart) to design and develop an automated track cleaning system. JAMBER Kart’s solution is a cart- mounted attachment that uses a sophisticated vacuum head to vacuum any debris, while simultaneously pumping isopropyl alcohol (IPA) and wiping down the inner surfaces of the track. A sealed and protected 3D-printed enclosure houses all of the electrical components and filters out collected debris, keeping Texas Instruments’ facility clean and ready for production at all times.

Phoenix Engineering Consultants were tasked with optimizing Strukmyer Medical's manufacturing process for circular adhesive patches bonded to irregularly shaped substrates. The labor-intensive manual stacking process caused high labor costs. Our solution automates this process by organizing patches into neat rows directly from the rotary die cutter, using a sloped vibration plate and conveyor belt. These rows are then funneled into magazines, mimicking the manual stacking configuration. This automation reduces labor costs, improves process efficiency, and increases reliability by minimizing part failure risk. The new system can potentially save Strukmyer Medical up to $0.50 in labor cost per patch. By automating the patch organization process, we eliminated the bottleneck and significantly reduced labor requirements, enhancing overall efficiency and sustainability in the production line. This project represents a significant advancement in Strukmyer Medical’s manufacturing operations, ensuring consistent performance and substantial cost savings.

The purpose of this project is to address the need of an automatic label applicator for cylindrical containers to enhance START International’s product portfolio. While the current (manual) label applicator LAB01 meets basic market demands, more and more applications require an enhanced product that allows application of one or two labels, application on tapered containers, and faster application time, among other requirements. To meet these requirements, our team has developed an automated bottle labeling machine equipped with motorized rollers and a gap sensor system. This system enables precise application of one or two labels, with a desired gap between the two labels, on cylindrical, tapered, and tabbed containers of varying sizes. By minimizing manual intervention and reducing labeling inconsistencies, our solution significantly improves efficiency and meets industry demands for a flexible, high-quality labeling solution. The automated machine streamlines operations and boosts productivity, making it the perfect addition to START International’s Portfolio.

Our client, Sleep Technologies, have developed and created a “Smart Bed”, comprised of smart rail technology and real-time computer analysis to adapt and adjust to an individual using the bed. Due to its complex and delicate componentry; special care needs to be taken when it comes to the shipping of the bed. As such, we have been tasked with designing an all-encompassing, holistic packaging and delivery system for the shipment of the bed. We achieved this by carefully specifying the delivery van, ordering custom made corrugated plastic boxes, as well as military grade cases that will be holding the components. A spaciously optimized van layout and loading/unloading process was made as well to allow for quick, simple, and easy delivery. Our solution not only increased protection for valuable components and overall transportation efficiency, it also gave the process a much more professional, sleek, and elegant presentation.