Doctoral researcher · University of Waterloo

I build small,
working worlds.

I make complex systems tangible, testable and understandable: a digital twin of a greenhouse for my PhD, and apps people actually use, from a calm paper reader to a plank timer you steer with your face.

See my research See my apps ↗ Download CV ↓

Tap the 🍅 up top for a plant fun fact. Shift-click it for tomato rain.

01 / Research

Computational research for greenhouse systems.

My PhD connects sensing, crop physiology and control in one modelling loop, so a greenhouse can be simulated, diagnosed and steered.

Research · public summary

A working twin of a real greenhouse

Greenhouses are partially observable, biologically variable and safety-critical, which makes them a natural testbed for state estimation, feedback and digital-twin methods.

  • ApproachCouple greenhouse sensing with crop-growth models and control, so the system can be simulated, diagnosed and steered.
  • SupervisorsProf. Nasser Mohieddin Abukhdeir and Prof. Christian Euler, Chemical Engineering
  • DisclosureMethods, data and results stay private until publication and IP review clear them. A fuller case study will appear here then.

Research themes

  • Scientific computing
  • Modelling
  • Sensing
  • Machine learning
  • Digital twins

A plant fact a day

Verified facts about plants, fruit and the odd corners of my research, with a new one every day.

Live figure

Sugar, from leaf to fruit

Leaves turn light into sugar, the upper ones shade the rest, and the stem carries it to the fruit that pull hardest. At night, starch banked by day keeps it moving.

Light0%Sugar reaching fruitlast 60 s

Tomatoes picked0tap a red one

A hand-tuned sketch of the flow my PhD models, not the model itself.

02 / Shipped

Apps I've shipped.

Designed by me, built with friends and AI. Each one started as a problem in my own day.

Phloem: Paper Reader

A calm reading desk for research papers on iPad

On the App Store

A reading guide that walks you down the page, real page turns, and Apple Pencil notes beside the source. New on the web: faster Workspace handwriting, sticky notes on PDF pages, and an export that writes your marks into the PDF. AI is optional and privacy-first.

Why I made itMy professor told me to stop letting AI summarize papers and actually read them. No reader I tried kept me on the page, so I built one. I named it after my PhD: phloem carries sugar from the leaves to the fruit, and a reader carries knowledge from paper to brain. Say it fast and it's "flow them", which is kind of the point.

Download free on theApp Store Open the web reader →

Life Game Project

Four games for iPhone and iPad, and a browser arcade

Game Life and Qi on the App Store · two more coming soon

Conway's universe you can hear, Go that teaches as you play, a plank timer you steer with your face, and a Brazilian card game for friends.

Game LifeApp Store QiApp Store Plank Constantbeta Manilhacoming soon

Why I made themSomething to stare at while I plank, and games to play with the girl I liked, who later ghosted me (or the situation became too awkward to talk, idk, idc). The games stayed.

Visit lifegameproject.com ↗ Play in your browser

03 / Selected work

Proof lives in the prototype.

A method matters when it survives contact with reality and leaves its evidence visible.

Routing · open notebook

Smart evacuation maps

Problem
Shortest-path routing can pile crowds into the same exits during large evacuations.
Method
Dijkstra routing, ant-colony-inspired optimization and LLM-assisted decisions in one agent-based framework.
Evidence
Outperformed plain Dijkstra in the tested majority-evacuation scenarios.
GitHub ↗ Build note ↗

Physical computing · light

S.W.I.M., the Sequential Wave Imprinting Machine

Idea
Move a one-dimensional strip of light through space to paint a two-dimensional image.
Build
Arduino, addressable LEDs and long-exposure photography turn digital pixels into a physical trace.
Context
A hardware project sponsored by Prof. Steve Mann.
GitHub ↗

04 / Experience

Research is the thread. Range is the toolkit.

Teaching, product engineering and consulting give me the range to carry ideas into real systems.

01Research

  • PhD Researcher, Chemical EngineeringUniversity of Waterloo2026 – now

    Computational modelling, sensing and machine learning for greenhouse systems.

  • Research ContributorJilin Agricultural Science and Technology University2025

    Analyzed multi-year field-trial data (soil chemistry, spectroscopy, FTIR, crop yield) with Python and PCA, and supported the manuscript.

  • Research AssistantProf. Yuning Li's lab, University of Waterloo2019

    Chromatography and UV-vis spectroscopy for polymer semiconductor and organic thin-film research.

02Teaching

  • Teaching Assistant, Machine Learning & PythonUniversity of Toronto2024 – 2025

    APS1070 Foundations of Data Analytics and ML, CSC108 Introduction to Programming, and MIE370 Introduction to ML: tutorials, office hours, projects and exams for courses of up to 500+ students.

03Education

  • MEng, Electrical and Computer EngineeringUniversity of Toronto · Data Analytics and ML emphasis2023 – 2025

    Featured alumnus on the program page ↗

  • BASc, Nanotechnology Engineering (Co-op)University of Waterloo · Honours2018 – 2023

    President's Scholarship of Honour · President's Research Award

04Industry

  • WiFi Software Testing Engineer (Intern)Ford Motor Company of Canada2022

    Benchmarked SYNC and TCU connectivity with iPerf3 across TCP, UDP, FTP and HTTP; built Ubuntu client-server test rigs.

  • QA Automation (Co-op)Manulife · Teranet · Imagine Communications2020 – 2021

    Selenium regression, UI and monitoring tests in JavaScript and C# across three four-month co-op terms.

05Consulting

  • Complaints Management & Internal Consulting (Intern)FAW-Volkswagen2025

    Investigated high-risk complaints across dealerships and teams, and built Kanban dashboards to track cases and recurring issues.

06Credentials

  • Machine Learning SpecializationStanford University & DeepLearning.AI2025
  • SQL for Data Science · Algorithms and Data StructuresUC Davis · Stanford Online2025

Languages

Python, C++, JavaScript, Swift, Java, C#, R, SQL

Machine learning

PyTorch, TensorFlow, scikit-learn, PCA, regression, classification, clustering, MediaPipe

Data & modelling

pandas, NumPy, SciPy, COBRApy, SBML, linear programming, sensitivity analysis, calibration

Tools

SwiftUI, Git, Docker, Linux, AWS Lambda, REST APIs, Selenium

05 / Desk tools

Small tools, private by default.

Things I built for my own desk. They run in your browser and keep your data on your device.

UnclenchA camera game for gentle neck and shoulder breaks Monet LightPaint your own photo in light, in about a minute The UnderstoryNatural facts, hidden questions JournalA quiet, private daily log Note boxLeave me a note, no login Finance trackerLocal-first, imports OFX/QFX Screenshot StudioCrop screenshots to exact App Store sizes, locally Phloem (web)The reader, in any browser

06 / Contact

Let's make something useful.

I'm open to research collaborations, technical conversations and ambitious systems problems, especially where modelling, sensing and machine learning overlap.

Email Houfu