Goal: Learn to estimate, simplify, and reason under uncertainty—fast.
Resources:
- Art of Approximation & Street-Fighting Math (read selectively, do examples) — dimensional analysis, extreme cases, scaling, back-of-envelope
- Information & Entropy (MIT 6.050J) — bits, noise, compression; bridges nicely to signals and experiment limits. Skim Units 1–4 & 8 (inference)
Link: https://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-050j-information-and-entropy-spring-2008/
Lablet: Do 3 'Fermi' estimates for your setup (e.g., expected electrode polarization drift vs time; SNR before/after salt mist). Write one page of assumptions and a numeric range for each.
https://engineering.purdue.edu/~ce474/Docs/Street-fighting%20mathematics.pdf
Goal: Speak the language of time series and filters so your pipeline choices are principled.
Resources:
- Signals & Systems (MIT 6.003) — sampling, Fourier/Laplace, LTI intuition (watch lectures + do a few problem sets)
Link: https://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-003-signals-and-systems-fall-2011/
- Discrete-Time Signal Processing (MIT 6.341) — windowing, FFT/DFT, filter design, STFT; skim notes and exam solutions to internalize patterns
Link: https://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-341-discrete-time-signal-processing-fall-2005/
- Optional supplement: RES.6-007 for analog/digital bridges, if you want more intuition
Lablet: Export a 1-hour raw channel from your ESP32 logs. Compute PSD and band-powers with multiple windows; compare Welch vs. simple FFT. Design a minimal FIR to reject mains + low-freq drift; document pass/stop bands and the trade-off.
Goal: Clean, stable measurements: grounding, shielding, impedance, ADC realities.
Resources:
- Stanford ENGR40M notes (charge, KVL/KCL, diodes/transistors) — skim to refresh circuit basics with a modern slant
Link: https://web.stanford.edu/class/engr40m/
- Analog Devices MT-series app notes for ADCs, filters, noise, and grounding (skim topic-wise as needed)
Link: https://www.analog.com/en/technical-articles.html
Lablet: Build a repeatable noise-audit: measure your system's baseline with shorted inputs, then with saline cup, then with electrodes open in air; plot histograms and Allan deviation. Record exactly what changes when you move grounds.
Goal: Understand what your electrodes are 'seeing,' and what EIS can (and can't) tell you.
Resources:
- MIT Electrochemical Energy Systems (10.626) — equivalent circuits, kinetics, transport; browse the lecture notes package
Link: https://ocw.mit.edu/courses/chemical-engineering/10-626-electrochemical-energy-systems-spring-2014/
- Bazant's electrokinetics notes for intuition on interfaces and double layers
- EIS fundamentals — Gamry application notes (Part I–IV), plus a concise 2023 ACS tutorial overview
Link: https://www.gamry.com/application-notes/EIS/
Lablet: Do a three-condition impedance check: needle pair in air, in RO water, in your saline mix. Sweep low-to-mid frequencies; fit a simple Randles-like circuit. Compare fitted Rs, Cdl, and a Warburg-ish element between conditions.
Goal: Soil as an electro-bio-physical system; what 'healthy' means and how to measure it.
Resources:
- Cornell Comprehensive Assessment of Soil Health (CASH) Manual (3rd Ed.) — indicators, procedures, interpretation
Link: https://www.css.cornell.edu/extension/soil-health/manual
- USDA NRCS Soil Biology Primer — soil food web overview, roles of fungi
Link: https://www.nrcs.usda.gov/resources/guides-and-instructions/soil-biology-primer
- Optional open text: Intro to Soil Science (Open Textbook) for fundamentals
Lablet: Build a soil-health checklist for any test plot you touch (texture, OM proxy, infiltration mini-test, simple EC, pH). Align each metric with a hypothesis about how it should shift your signal statistics.
Goal: Enough mycology to design sensible, controlled experiments with living substrates.
Resources:
- Intro resources (university course pages/notes or open texts) to cover morphology, growth, reproduction, and mycorrhizae
- UAlberta's mycology catalog page for scope
Link: https://sites.ualberta.ca/~ginns/mycol/
- APS 'Intro to Fungi'
Link: https://www.apsnet.org/edcenter/disimpactmngmnt/topc/Pages/IntrotoFungi.aspx
- Open mycology text
Lablet: Draft a strain + substrate protocol for two species (oyster + one AMF inoculum), capturing temperature/humidity, colonization stages, and a stabilization window before stimulation.
Goal: Treat your data as random processes; avoid p-value theater; focus on effect sizes and prediction.
Resources:
- MIT Probability (6.041/6.041SC) — random variables, transforms, simple processes, basics of inference (lecture videos or the SC site)
Link: https://ocw.mit.edu/courses/electrical-engineering-and-computer-science/6-041sc-probabilistic-systems-analysis-and-applied-probability-fall-2013/
- (Optional) Think DSP for practical Pythonic DSP habits (skim chapters on spectra, correlation)
Link: https://greenteapress.com/thinkdsp/
Lablet: Pre-register a mini-experiment: define a stimulus, a detection metric (e.g., band-power change, cross-correlation vs baseline), and a stopping rule. Run N=10 replicates; report confidence intervals and a predictive check (split-sample).
Goal: Change fewer things; learn faster.
Resources:
- Penn State DoE/Regression notes — encode simple randomized block designs; avoid confounding; power thinking
Link: https://online.stat.psu.edu/stat503/
- Pair with the Information & Entropy unit on inference for a unifying lens (MIT 6.050J)
Lablet: Plan a 2×2 (or Latin square) comparing two substrates × two electrode pre-treatments with randomization. Pre-compute required N for a target detectable effect on one metric you care about.
Goal: Keep the 'whole' in view: emergence, networks, error-correcting patterns vs noise.
Resources:
- SFI's Complexity Explorer — Introduction to Complexity (audit/skim modules on networks, information, emergence)
Link: https://www.complexityexplorer.org/courses/89-introduction-to-complexity
- Revisit Street-Fighting Math techniques to simplify models you're tempted to over-fit
Lablet: Re-express one detection pipeline in three simpler ways (e.g., threshold on log-band-power; simple AR(1) residuals; moving-median deviate) and see if decisions agree >80% of the time.
Goal: Work like a responsible lab; codify your protocols; make results repeatable.
Resources:
- WHO Laboratory Biosafety Manual (4th ed.) + CDC BMBL (6th ed.) — focus BSL-1 practices relevant to fungi and benign soil work
Link: https://www.who.int/publications/i/item/9789240011311
Link: https://www.cdc.gov/labs/BMBL.html
- (Optional) University biosafety manuals for concrete checklists
Capstone Lablet (Month-long): Lock one end-to-end protocol: setup → calibration → stimulus → recording → preprocessing → statistic → report. Include a one-page threats-to-validity section (temperature swings, electrode fouling, saline variability, cable movement), and a 10-item reproducibility checklist that another person could follow.