Che-Hao Hsu Email & Phone Number

Taiwan

Miaoli County, Taiwan

Redmond, Washington, United States

• Mentor: Brittany Wegmann, Prof. Denise WilsonWe aimed to build a capnography device to detect human vital signs, including end-tidal CO2 level, lung pressure, and flow rate of exhaled breath. I was assigned to the.
• Executed CO2 calibration with multiple CO2 concentrations to establish the modified Beer-Lambert equation for our system and performed the real-time measurement with 4% CO2 with an accuracy of approximately 92.28%.
• Enhanced eighteen times gain for the gas and reference signals by modifying the bandpass filter and amplifier circuit and ran experimental tests with an oscilloscope to validate the outcome.
• Troubleshot the cause for the incorrect CO2 measurements by breaking down the whole system into pieces to track and analyze voltage signals from the IR sensor, IR emitter, microcontroller, and circuits.

Taoyuan City, Taiwan

• Managed manufacturing flow and organized the component requirements of the product prototype, saving approximately 33% of production time.
• Developed five controllers for the communication between robotic arms and ultrasound sensors by soldering circuits and integrating the microcontroller, relay, and I/O modules.
• Manufactured EMI shielding on all electronic components and input/ output connections of controllers to prevent signals from being disturbed by one another or external electromagnetic noise.
• Validated and debugged controller prototypes by measuring the input/ output signals of electronic devices with oscilloscopes and running controllers with ultrasound sensors and robotic arms under practical situations.

Taoyuan City, Taiwan

• Advisor: Prof. Pei-Yun TsaiThe goal was to make an autonomous car that can find the fastest path to leave a maze in a deep reinforcement learning way. Our team attempted to build a reinforcement learning processor for.
• Created 72 multiply-accumulate (MAC) units, floating point arithmetic unit, and memory units for forward and backward propagation computation via RTL design in Verilog.
• Reduced 46.5% computation time by data forwarding and scheduling instructions to remove stalls and hazards.
• Built floating point arithmetic elements to perform high-precision computation for learning neural networks.

Master Of Science - Ms, Electrical And Computer Engineering, 3.82

Courses: Computer Architecture, Computer Systems Architecture, Digital System Design with FPGAs, Engineering Entrepreneurial Capstone.

Bachelor Of Science - Bs, Electrical Engineering, 3..8

Activities and Societies: Sports Officer in Student Association of NCUEE Camp Leader for NCU Voluntary Service Winter Camp General Affair.