Biofluid Analyzer

Mechanical Design and Iterations

Using the early sketches as a foundation, I translated the selected concept into a detailed CAD model in SolidWorks, defining the enclosure geometry, mounting features, and internal component layout. I made deliberate design decisions around how the device would interface with the bed rail, including orienting the load cell slot parallel to the bedside to reduce the risk of accidental contact during routine patient care. The CAD model served as the reference for dimensional tolerancing, component fit, and downstream prototyping, ensuring the design remained compact, nurse-accessible, and compatible with the clinical environment.

In parallel, I led the team through an independent design-and-build phase where each member developed their own bedside attachment concept, modeled it in CAD, and produced a 3D-printed prototype. We evaluated all prototypes directly in the hospital, testing for fit, stability, ergonomics, and ease of interaction with the bed rail and Foley bag under real use conditions. Based on hands-on testing and clinician feedback, we converged on the design I developed (printed in red PLA), which demonstrated the most secure attachment to the bedside and the most stable weight distribution when the Foley bag was loaded, making it the best candidate for continued iteration and clinical use.

Electrical Design

I led the electrical system design to support accurate bedside weight measurement, real-time volume conversion, and continuous user-facing display within a compact enclosure. The system was built around a 5 kg load cell paired with an HX711 instrumentation amplifier to capture high-resolution mass measurements, interfaced with an ESP8266 NodeMCU for computation and control. An I2C-based LCD display provided continuous bedside readout, while a dedicated push button enabled on-demand taring to support nurse workflows. The circuit was initially assembled on a breadboard using jumper wiring to enable rapid iteration and calibration before enclosure integration, with a transformer-based power supply selected to support stable operation in a clinical environment.

Code Architecture and Embedded Software

The embedded codebase was structured to support calibration, repeatable mass-to-volume conversion, and real-time bedside feedback using the HX711 sensing stack.

• Load cell calibration: Implemented a serial-based calibration routine using the HX711 library, allowing dynamic adjustment of the calibration factor and taring to establish an accurate baseline before clinical use.

• Mass acquisition: Sampled load cell readings using averaged measurements to reduce noise and improve repeatability across Foley bag weight ranges.

• Mass-to-volume conversion: Converted measured mass into urine volume using a calibrated density-based conversion, enabling direct volume readout in milliliters rather than raw weight values.

• Bedside display: Drove a 16×2 I2C LCD to present continuous, human-readable urine output at the bedside without requiring external devices.

Shipping the MVP

After completing bench validation, I led the delivery and in-hospital deployment of the MVP at UF Health Shands Hospital for bedside evaluation. I coordinated the handoff with the physician sponsor and nursing staff, ensuring the device was installed on active patient beds and used within the existing care workflow. During this phase, we observed real-world usage, gathered structured feedback from clinicians, and evaluated ergonomics, readability, and interaction under clinical conditions.

Clinical Feedback and Iteration

During in-hospital use across multiple units, real-world handling revealed durability limitations in the initial enclosure, particularly at the bed-rail attachment points where repeated mounting and removal introduced wear and localized stress. Based on this feedback, I led a targeted design iteration to reinforce high-load regions and reprint the enclosure using medical-grade PLA better suited for repeated clinical handling and cleaning. This iteration improved structural robustness while preserving the original form factor and bedside accessibility, ensuring the device could withstand sustained clinical use without compromising usability.

Clinical Deployment & Pilot Study

Following the durability iteration, I led the final delivery and installation of the device for active use within UF Health Shands Hospital. After observing its integration into daily nursing workflows, the physician sponsor initiated a formal pilot study to evaluate the device’s impact on the efficiency and accuracy of urine output data collection for long-term patients. I collaborated directly with the physician to outline study objectives and usage protocols, and we discussed pathways for expanding the device’s deployment across additional departments and units within the hospital. This phase marked the transition from prototype validation to clinical evaluation and adoption planning.

As part of the clinical deployment effort, I authored a comprehensive user manual to support onboarding of new clinicians using clear, customer-facing language. The documentation covered end-to-end device setup, operation, validation checks, and troubleshooting, enabling nursing staff to adopt the device consistently across shifts and units without direct engineering support. This ensured safe, repeatable use in a clinical environment and reduced friction during early adoption and pilot study rollout.