🔒 COPPA Compliant 🛡️ FERPA Ready 📋 NYS Ed Law 2-d 🌱 Hands-On First 🧪 Scientific Method

Students Grow Plants.
Students Do the Thinking.
Rowan Guides the Science.

A classroom garden STEM platform where students observe, measure, hypothesize, and reason — before the technology responds. Built on years of hands-on agricultural education with the One Planet Education Network.

Rowan Guardian Teacher Dashboard showing classroom garden management, student progress, and real-time sensor data

Teacher Dashboard — Real classroom garden platform. Now enrolling pilots for 2026–2027.

Why Cognosense AI Was Formed

Our team has been working on the ground with the One Planet Education Network (OPEN) for years, bringing hands-on agricultural science to classrooms worldwide — long before Cognosense AI, LLC was even formed.

During this work, we observed a terrifying, highly documented trend: generative, conversational tools are actively destroying the critical thinking, writing, and analytical skills of the students who use them. Rather than wrestling with problems, students are outsourcing their minds to screens — a dangerous process researchers call "cognitive off-loading."

We formed Cognosense AI, LLC specifically to combat this crisis. We built Rowan Guardian Classroom to act as the antidote: a strictly controlled, closed-loop Socratic Digital Mentor that is software-locked to require physical interaction with sensors, soil, and real plant observation before any guidance is unlocked.

This is a brand-new software launch. We are currently enrolling our first cohort of school pilots for the 2026–2027 academic year. We are starting small, building deep trust, and working with every pilot school as true partners.

We Built Rowan to Address the Concerns You Already Have

Parents, teachers, and researchers are raising important questions about technology in schools. We listened before we built.

🧠

The Concern: "Technology does the thinking for students"

Rowan never gives a direct answer first. It asks students to observe their plant, touch the soil, check their sensors, and form a hypothesis — then discusses what the data suggests. The student does the thinking. Rowan guides the process.

This is the Socratic method — the same teaching approach used for 2,400 years.

🚫

The Concern: "Children form bonds with chatbot personalities"

Rowan is not a character, companion, or friend. It has no persona, no backstory, no emotional simulation. It is a structured science notebook — a lab journal that guides students through observation, hypothesis, and evidence. The relationship is with the plant and the data, not with the software.

Students interact with a scientific process, not a personality.

🌱

The Concern: "Little evidence these tools improve learning"

Rowan doesn't try to replace instruction. Students grow real plants, collect real sensor data, make real observations, and write structured science journals. The learning comes from the hands-on activity — the technology supports the scientific method, it doesn't substitute for it.

We Agree. Students Shouldn't Outsource Their Thinking.

Here's the difference between a generic chatbot and Rowan's Socratic approach — in a real student interaction.

❌ Typical Generative Chatbot

Student: "How much water does my bean plant need?"
Chatbot: "Bean plants typically need 1–2 inches of water per week. Water deeply when the top inch of soil is dry. Avoid overwatering as this can lead to root rot."
✗ Student reads answer. Student learns nothing. Brain is OFF.

✅ Rowan Guardian — Think First Protocol™

Student: "How much water does my bean plant need?"
Mentor: "Great question! Let's figure it out together. 🌱

First — is your plant in a big pot or a small one? And does the soil feel dry when you touch it?

Try poking your finger in the soil up to your first knuckle. What does it feel like?"

⭐ +10 XP for asking a great question!
Student: "The soil feels kind of dry..."
Mentor: "Good observation! If the soil is dry and the pot is small, what do you think that means for how often you should water?"
  • Student observes (looks at the pot)
  • Student collects data (touches the soil)
  • Student hypothesizes (connects pot size to water needs)
  • Student practices the scientific method. Brain is ON.

This isn't a chatbot in a garden costume.
It's the scientific method, made interactive.

Built on Evidence, Not Hype

We read the same research that parents and educators are reading. Here's how it shaped Rowan's design.

Stanford SCALE Review (2026)
"Tools with pedagogical guardrails — such as those that provide step-by-step reasoning rather than direct answers — show more promise than general-purpose tools."
How this shaped Rowan: The Think First Protocol™ is a 6-stage structured reasoning framework. Students never receive a shortcut answer — they follow a rigorous reasoning process that builds critical thinking skills.
Stanford Accelerator for Learning (SCALE), 2026
Dr. Jared Cooney Horvath (2026)
"The tools an expert uses to simplify their work are not appropriate for a student trying to develop mastery."
How this shaped Rowan: We designed a LEARNING tool, not an efficiency tool. Every interaction requires the student to observe, hypothesize, and evaluate BEFORE receiving guidance. Cognitive effort is the point, not a bug.
Dr. J.C. Horvath, The Guardian, June 2026
AFT "Hands-On" Position (May 2026)
"Devices down, eyes up, hands-on." Remove student-facing generative tools from elementary schools. Technology should support teachers, not replace them.
How this shaped Rowan: Rowan IS hands-on. Students grow real plants, touch real soil, measure real temperature and humidity. The teacher controls every aspect of the student experience. Rowan extends the teacher's reach — it never replaces their judgment.
American Federation of Teachers, May 27, 2026

The 6 Stages of the Think First Protocol™

Every diagnostic analysis runs through our strict, 6-stage reasoning engine before the Socratic mentor responds — ensuring absolute data integrity and leaving the scientific decision to the student.

1

The Framing Brief

Defines the scientific objective, constraints, and school-safe parameters of the inquiry before any reasoning begins.

🧠 Skill: Precise Problem Definition
2

Context Fusion

Merges raw physical sensor readings (soil moisture, pH, temperature), crop history, meteorological alerts, and regional soil profiles into a unified, non-personal data payload.

📡 Skill: Data Gathering & Cross-Referencing
3

Assumption Hunting (The Pre-Mortem)

Actively hunts for and exposes hidden biases, sensor errors, or weak links in our telemetry data before any hypothesis is formed.

🔎 Skill: Critical Evaluation & Bias Detection
4

Multi-Angle Analysis (The Red Team)

Forces the reasoning engine to analyze the problem from skeptical, adversarial, and competing horticultural perspectives. "What if it's NOT nitrogen deficiency?"

⚔️ Skill: Hypothesis Generation & Challenge
5

Falsification Check

Stress-tests the initial hypothesis against our local database of peer-reviewed botanical facts (derived from university extensions like Cornell, USDA, NOAA) to prove why the conclusion might be wrong.

🧪 Skill: Falsification & Evidence Testing
6

Synthesis & Decision (The Confidence Statement)

Constructs a final, evidence-tagged Socratic response with a mathematically verified confidence score (e.g., "Confidence: 95%") — putting the student back in the driver's seat.

📊 Skill: Evidence Synthesis & Confidence Assessment
Rowan Guardian Student Dashboard showing plant observations, sensor data, and guided scientific inquiry

Student Dashboard — Guided scientific inquiry with real plant data.

Other tools skip all 6 stages and go straight to the answer.
Rowan requires the student to engage with the full reasoning process before any guidance is provided.

What Happens When a Student Asks a Question

Every student interaction passes through four engineering controls — not policies that can be circumvented, but software-enforced safeguards.

🧹

PII Scanner

Bidirectional scanner checks for accidentally shared personal info (email, phone, address) and removes it before processing.

🚧

Subject Guardrails

Is this about STEM? Soil, plants, sensors, botany? ✅ Proceed. Anything else? Gentle redirect: "Let's look back at our garden sensors."

🔬

6-Stage Think First Protocol™

Every inquiry runs through all 6 stages — Framing, Context Fusion, Assumption Hunting, Red Team, Falsification, and Synthesis — before any response is shown.

Output Verification

Response verified against peer-reviewed sources (USDA, NOAA, Cornell Extension). Age-appropriate language filter applied.

What Students Cannot Do

These are engineering controls baked into the software — not policies a student can work around.

🚫 Cannot ask off-topic questions (subject guardrails)
🚫 Cannot communicate with other users
🚫 Cannot share data outside their classroom
🚫 No student audio is ever recorded
🚫 No student PII collected (first name only)
🚫 No generative creative content — structured STEM responses only
🚫 No advertising, no tracking pixels, no third-party cookies
🚫 No student data used for model training

Not a Chatbot. A Science Journal.

Students don't "chat" with Rowan. They record structured observations in a diagnostic science journal — just like a lab notebook. Every entry runs through all 6 stages of the Think First Protocol™.

🔬 Science Journal — Diagnostic Entry #12 March 15 · Student: AJ
Observation
"My tomato plant's lower leaves are turning yellow. The upper leaves look fine."
Stage 1 · Framing Brief
Objective: Diagnose cause of lower-leaf chlorosis on classroom cherry tomato (Week 6).
Stage 2 · Context Fusion
Sensor data: Soil moisture 48% · EC 1.2 mS/cm · Temp 24°C · Last feeding: 5 days ago · Weather: partly cloudy, no frost risk.
Stage 3 · Assumption Hunt
⚠️ Sensor check: Moisture probe last calibrated 10 days ago — readings within normal variance. No data flags.
Stage 4 · Multi-Angle Analysis
The Mentor prompts: "Good hypothesis! But if it were nitrogen deficiency, we'd expect ALL leaves to yellow — not just the bottom ones. And your EC reading of 1.2 tells us nitrogen is adequate. What else could explain bottom-only yellowing?"
Stage 5 · Falsification
Cross-referenced against Cornell Extension: "Bottom-leaf chlorosis with healthy upper canopy and adequate EC is consistent with natural senescence — not nutrient deficiency."
Stage 6 · Synthesis & Decision
Fact Bottom-only yellowing + adequate EC = natural senescence (normal aging).
Inference Remove yellowed leaves at the stem. Monitor upper leaves for 3 days.
Confidence: 92%
⭐ +25 XP for completing a full diagnostic · Badge: "Root Cause Finder" 🏆

What AJ Learned Today

  • Bottom-only yellowing = natural aging, not nutrient deficiency
  • EC readings tell you about nutrient levels in soil
  • The PATTERN of yellowing tells you the CAUSE
  • Always check your hypothesis against data before concluding
  • Assumptions can be checked — sensors can have errors too

AJ will recognize this pattern all season. Next time tomato leaves yellow, AJ won't need Rowan — AJ will know to check whether it's bottom-only, check the EC, and diagnose it independently.

Evaluating this for your school?

This page is designed to answer the questions your school board, PTA, or district administration will ask. Share it directly.

📤 Share With Your School Board

You Teach. Your Garden Grows. The Data Proves It.

Teachers control everything. Rowan is a tool that extends your reach — never replaces your judgment.

📚

Curriculum Courses

Generate NGSS-aligned plant science courses with structured modules, quizzes, and hands-on activities. Adapt to any grade level.

👁️

Full Visibility

See every student interaction, observation, and diagnostic entry. Nothing happens privately between a student and the tool.

📊

Student Progress

Track XP, completed observations, diagnostic accuracy, and scientific reasoning development across your entire class.

Rowan Guardian Teacher Tools showing course creation, student management, and classroom analytics

Teacher Tools — Course creation, student management, and progress tracking.

Built by Trusted Partners

Cognosense

Cognosense AI, LLC

Development Team & Platform Architects

Founded specifically to combat cognitive off-loading in education. Developers of the 6-Stage Think First Protocol™, the Rowan Guardian platform, and the complete data security infrastructure. The team behind the technology.

One Planet Education Network

One Planet Education Network (OPEN)

Education & Agriculture Partner

Years of hands-on agricultural education and curriculum delivery in classrooms and communities worldwide. OPEN provides the field-tested educational framework and global network that Rowan Guardian is built upon.

Rowan Guardian is built by Cognosense AI, in partnership with OPEN. Cognosense leads the technology platform, data security, and the 6-Stage Think First Protocol™. OPEN provides educational expertise and global agricultural networks.

Join Our Early Adopter Pilot

This is a brand-new launch. We are enrolling a limited number of schools for deep, hands-on pilot integration. Every pilot school works directly with our team. Your feedback shapes the product.

We are in our first year. We are measuring learning outcomes rigorously. If we find our approach doesn't produce the results we expect, we'll publish that too.

Questions? Reach us at pilot@rowanguardian.com