Machine learning for messy, real-world data.

• Current work · ongoing research

  $1.5M project funding

  Simulated data · for illustration

  Crown 3 2,477°F Crown thermocouple

  Underglass 1 2,184°F Underglass · doghouse end

  Underglass 2 sensor lost Underglass · mid-melt

  Underglass 3 2,126°F Underglass · throat end

  Virtual sensor · underglass 2 2,155°F Carrying the lost reading

  Glass furnace · The University of Toledo · NOIC industry project

  The sensor fails. The reading continues.

  Thermocouples in molten glass drift, degrade and fail — and they measure the variable that sets fiber diameter, breakage and energy cost. Six gradient-boosted models read underglass temperature from the crown instead, at 4.37°F RMSE mid-melt at 100 TPD.

  Read the case study →

• $500K NIH-funded research Cell death · The University of Toledo

  Reading cell death without a stain.

  Studying how cancer cells die normally requires fluorescent labeling, which is invasive and can alter the biology being observed. This asks whether death patterns can be recognized from raw, unlabeled microscopy video alone.
• 88% approx., identifying at-risk patients Osteoporosis · MS thesis research, The University of Toledo

  Risk, from a scan you already had.

  I isolate bone from a routine CT, extract 18 density features and combine them with age, BMI and lab markers — high, medium or low risk in minutes, at approximately 88% accuracy. It reads imaging patients have already had, instead of the separate DEXA appointment screening normally depends on.
• YOLOv8 + DeepSORT Real-time on a Raspberry Pi · ODOT statewide network Speed estimation · The University of Toledo · ODOT-funded

  Speed, from a camera that only sees boxes.

  Built first independently — roadside video capture, detection, tracking and speed derived from tracked movement, with real-time inference on a Raspberry Pi at the roadside. Then integrated into ODOT’s existing statewide detection network.
• 97.87% peak accuracy, hybrid model · undergraduate research Brain tumor · The University of Toledo, Dept. of EECS

  Where the boundary is not a boundary.

  Tumor margins are soft and blurred, which is exactly where hard clustering fails. Fuzzy C-Means segments the ambiguous boundary and a CNN classifies what it finds.
• Patent granted Indian patent IN 573184 FlowerBot · IEEE ICECA 2022 · Indian patent IN 573184

  FlowerBot — flower-plucking robotic arm

  Rose harvesting is manual, painful and impossible after dark: thorn injuries, sustained bending, no visibility at night. A PIC-controlled arm with night-vision cameras distinguishes flowers from leaves and bushes, detects obstacles, and slices and holds the flower in one motion, wet or dry.
• 8.92% MAPE across 919 count stations Traffic forecasting · The University of Toledo · ODOT-funded

  Seven years of counts is not a time series.

  With only seven years of counts, time-series forecasting is the wrong frame. Normalizing the origin-destination matrix by population turns the task into learning a spatial assignment operator that can be re-expanded against any year.

Also SmartPark — AI parking space detection · Senior design project, The University of Toledo / Security Discovery Tool API · Park Place Technologies / Pavement Intelligence (Power BI) · The University of Toledo · ODOT-funded project

• The osteoporosis work, written upPress · September 2026
• OTA 2026, from the exhibit hallConference · September 2026
• Graduation, and on to the master’sGraduation · May 2026