Project FLOW at Eagle Valley Middle School
Middle schoolers as practicing watershed scientists on the Carson River
October 4, 2026
The problem
Students rarely touch real environmental data. Classroom kits produce numbers nobody downstream can use, and the professional software that handles real sensor data is built for agencies, not twelve-year-olds. Project FLOW set out to close that gap: students collecting real data on their own river and running it through the same workflow a professional would.
The data
- Field measurements on the Carson River at Riverview Park: water temperature, specific conductance, and dissolved oxygen, first from continuous loggers and then by hand one to two times a month.
- Field bursts: short logger recordings taken while holding the meter, so a class can watch the sensors settle before choosing the value for each parameter.
- Context pulled live from public sources (USGS streamflow, SNOTEL snowpack, NASA POWER weather, the Water Quality Portal, and EPA’s ATTAINS) so students can compare their numbers with the agencies'.
What I built
The Project FLOW app, a classroom configuration of the DANalytics engine:
- The professional workflow, in order: import, context, corrections, quality checks, clean, analysis, export. Every tab is a numbered step with a directive telling students what to do there, and the numbering is computed, so it stays right as features switch on and off.
- Bursts become field readings: a burst keeps its continuous shape so students can see the sensors equilibrate; the chosen value then enters the pipeline as one discrete reading.
- Nothing is ever deleted: quality checks flag; they never remove. Each flag records whether an algorithm, a person, or both raised it.
- A Data Quality Report Card that grades the record, not the students: how much survived review, how well the automatic and human flags agreed, how far the cleaning moved the median. It is designed to teach students to receive a critique of data they collected themselves.
- Built-in teaching: a single-source help system rendered as per-tab help, a guide, and a printable field manual, plus narrative lessons tied to what is on screen.
- Class reports rendered with Quarto in a background process, so one class generating a report never freezes another.
The chapter itself is half the work: a school, a lead educator, community partners, and a model other schools can adopt.
Tools
- R and Shiny (modules,
bslib) on a shared engine of pure functions, so the classroom app and the professional one never diverge. - Quarto for generated class and comparison reports.
callrwith Shiny’sExtendedTaskto run reports and public-data fetches in child processes with timeouts.testthat, with the whole suite run by GitHub Actions on every pull request.- shinyapps.io for the public demo, deployed by a script that pins the package versions from
renv.lock.
Outcome
- Launched April 30, 2026, with students placing equipment in the Carson River.
- Funded by a $5,000 Rotary Club of Carson City grant, with technical guidance and a donated temperature logger from the Nevada Division of Environmental Protection.
- When the field equipment was stolen in summer 2026, the chapter switched to hand-collected discrete measurements, which proved cheaper, harder to steal, and better at teaching where a number comes from. Carson City Parks, Recreation and Open Space joined on site design and a permanent interpretive sign.