Facilitator Master Agenda

CRAFT PD Series: Unlocking AI & IoT for K-12 STEM

Three-workshop master run-of-show for facilitators
Hosted by: UCF DRACO Lab (Dept. of ECE) & School of Teacher Education (CCIE)
Facilitators: Dr. Mike Borowczak & Dr. Andrea Borowczak
Series: Western Regional Noyce Network (WRNN)
Participants: ~20-25 Noyce recipients per session
Platform: Zoom (virtual, synchronous) — see links below
← Series Home ← Facilitator Resources W1 Agenda W2 Agenda W3 Agenda Polls Appendix ↓ Checklist ↓

Links — Quick Reference

Copy-paste these into Zoom chat at the appropriate times:

Workshop Zoom Pre-Survey Shared Doc Live Agenda Post-Survey
W1: Apr 11 Join Zoom Pre-Survey Shared Doc — Post-Survey
W2: Apr 18 Join Zoom Pre-Survey Shared Doc Live Agenda Post-Survey
W3: Apr 25 Join Zoom Pre-Survey Shared Doc Live Agenda Post-Survey

CRAFT Cycle Phases (All Workshops)

C
Contextualize: Where are participants starting? Activate prior knowledge.
R
Reframe: Shift perspective. Address fears. Build buy-in.
A
Assemble: Hands-on building. I Do → We Do → You Do scaffolding.
F
Fortify: Verify, test, check. Apply Check the Machine (CtM) protocol.
T
Transfer: Design something for YOUR classroom. Make it real.
Workshop 1 of 3

Workshop 1: Using AI for STEM

Date: April 11, 2026 | 8:30 AM – 12:00 PM PST
Zoom: https://ucf.zoom.us/j/96520248667?pwd=gqfMJJAKQ5fEOWTXrabpeIiTYuom0u.1
Focus: Curriculum alignment, differentiation, administrative time-saving
Talk:Do Ratio: 28 min facilitator / 162 min participant (1:5.8)
"Stop asking what is AI and start asking how does this help my students learn."
📝 Pre-Survey (share at 8:30) 📋 Shared Group Workspace (share at 8:30) 📊 Post-Survey (share at 11:40)

Session Materials (share as needed)

Facilitator-led (28 min) Participant activities (162 min) Breaks (20 min)
8:30
Admin 10 min
Welcome + Icebreaker + Pre-Survey
Welcome participants, quick introductions. Icebreaker: One word that describes your relationship with AI (type in chat). Pre-Survey: Open Pre-Survey. Shared Doc: Open Group Workspace. Facilitator: Share screen with survey QR code.
8:40
Listen Contextualize 15 min
What Is AI, Really?
Brief demo: same prompt on ChatGPT, Claude, Gemini side by side. Frame: AI is a tool, not magic. Show the range of outputs. Participants follow along on their own devices.
8:55
Do Reframe 20 min
AI as Co-Pilot, Not Autopilot
Poll: What's your biggest fear about AI in teaching? Breakout rooms (4-5 per group, 10 min): Discuss fears, then reframe as opportunities. Report-out: Each group shares their reframe (2 min each). Participants type takeaways in Shared Group Workspace.
9:15
Break 10 min
Break #1
Stretch, refill, remind everyone that W3 is simulator-only — kits ship afterwards, not before.
9:25
Do Assemble 25 min
Cross-Platform Prompt Engineering
I Do (5 min): Demo prompt engineering — show a bad prompt vs. good prompt. We Do (10 min): Groups test SAME prompt on 3 platforms, fill comparison table in Shared Doc. You Do (10 min): Each person writes their own prompt and tests it.
9:50
Do Assemble 25 min
Prompt Iteration Lab
Take a weak output, iterate on the prompt, document before/after. Groups work in their color-coded section of the Shared Doc. Facilitator circulates through breakout rooms.
10:15
Do Assemble 30 min
NGSS-Aligned Lesson Builder
Use preferred LLM to generate a lesson aligned to a specific NGSS standard. Customize for their grade band and subject. Add best prompts to the Collective Prompt Library table. Walk around, help with prompt specificity.
10:45
Break 10 min
Break #2
Stretch. Remind participants to add to Collective Prompt Library.
10:55
Do Fortify 20 min
Spot the Difference
Present two lessons (one AI, one human). Groups analyze in Shared Doc. Gallery walk: groups review each other's analyses. This plants the seed for Workshop 2's verification focus.
11:15
Do Transfer 25 min
Lesson Idea Board + CRAFT Debrief
Everyone adds to the Shared Lesson Idea Board. Brief CRAFT debrief: reveal the C→R→A→F→T structure they just experienced. "You are now an AI-equipped STEM educator. Go break things and learn from them."
11:40
Admin 20 min
Resources + Post-Survey + Close
Walk through Digital Toolkit. Post-Survey: Open Post-Survey. Remind: Workshop 2 in one week — we'll learn to verify AI outputs.
Workshop 2 of 3

Workshop 2: Verifying STEM AI Outputs

Date: April 18, 2026 | 8:30 AM – 12:00 PM PST
Zoom: https://ucf.zoom.us/j/96789217432?pwd=pzJPBMpRqaTlq3Hi5qe8NAexHXYTUd.1
Focus: Academic integrity, error analysis, critical thinking in STEM
Talk:Do Ratio: 26 min facilitator / 164 min participant (1:6.3)
"Your students are using AI — but do they know when it's wrong?"
📝 Pre-Survey (share at 8:30) 📋 Shared Group Workspace (share at 8:30) 🗓 Live Agenda (keep open on your side) 📊 Post-Survey (share at 11:40)

Session Materials (share as needed)

Facilitator-led (26 min) Participant activities (164 min) Breaks (20 min)
8:30
Admin 10 min
Welcome + Pre-Survey
Welcome back! Quick check-in: What did you try since Workshop 1? Pre-Survey: Open Pre-Survey. Shared Doc: Open Group Workspace.
8:40
Listen Contextualize 15 min
When AI Gets It Wrong
Live demo: Ask an LLM a STEM question it gets subtly wrong. Show real examples of AI hallucinations in code and science content. Dramatic reveal — participants don't know the errors until told.
8:55
Do Reframe 20 min
The Problem Isn't AI Use — It's Unverified AI Use
Poll: How confident are you at catching AI errors? (chat vote). Breakout discussion: When have students submitted unverified AI work? Groups capture insights in Shared Group Workspace.
9:15
Break 10 min
Break #1
Stretch. Tease the error hunt coming up.
9:25
Do Assemble Part A 30 min
Code Error Hunt
I Do (5 min): Walk through one buggy code snippet, model the debugging process. You Do (25 min): Breakout groups receive AI-generated code snippets with deliberate bugs. Groups compete: most verified bugs wins! Post findings in Shared Doc. Code snippets include: projectile calc (deg/rad), list sorting, unit conversion.
9:55
Do Assemble Part B 30 min
Content Error Hunt
Same format, now with AI-generated science/math explanations. Groups find factual errors and rate subtlety (Obvious / Tricky / Sneaky). Everyone adds discoveries to the Collective AI Error Gallery.
10:25
Break 10 min
Break #2
Stretch. Groups check their error counts.
10:35
Do Fortify 25 min
Check the Machine (CtM) Protocol Deep Dive
Introduce CtM: Task → Before → After → Takeaway. Each group applies CtM to a new AI output, posts full walkthrough in Shared Doc. Then: design a CtM activity for their own classroom (subject + grade). CtM is the key artifact from this workshop.
11:00
Do Transfer 20 min
AI Policy Committee
Groups draft verification-based classroom norms for AI use. Each group presents their top 2 norms. Compile best into shared template. "Your students don't need to be AI experts. They need to be AI skeptics."
11:20
Admin 20 min
Debrief + Resources + Post-Survey + Close
Debrief wall in Shared Doc. Share CtM template and Error Gallery links. Post-Survey: Open Post-Survey. Remind: Workshop 3 next week is simulator-only — no hardware needed to join. Kits ship AFTER W3 to participants who complete it.
Workshop 3 of 3

Workshop 3: Programming Edge/IoT Systems with AI

Date: April 25, 2026 | 8:30 AM – 12:00 PM PST
Zoom: https://ucf.zoom.us/j/97106116299?pwd=UojbRbbHx76prSljjq4rh4EdQvf5GK.1
Focus: Two-node IoT pattern (Sensor Node ↔ Aggregator Node) built entirely in the MakeCode online simulator. NGSS/CTE alignment, radio networking, LLM-assisted coding.
Hardware: NONE during the session — all activities run in makecode.microbit.org. BBC micro:bit V2 kits ship AFTER the session to participants who complete both surveys + submit a draft IoT lesson.
Talk:Do Ratio: 25 min facilitator / 165 min participant (1:6.6)
"IoT is a conversation between devices — and you can have that conversation in a browser tab."
📝 Pre-Survey (share at 8:30) 📋 Shared Group Workspace (share at 8:30) 📊 Post-Survey (share at 11:40)

Session Materials (share as needed)

Facilitator-led (25 min) Participant activities (165 min) Breaks (20 min)
8:30
Admin 10 min
Welcome + Simulator Check + Pre-Survey
Welcome! Ask everyone to open makecode.microbit.org in a browser and drop a show string "hi" in on start — confirm the simulator renders and scrolls. Demo that adding one radio.sendNumber block spawns a second simulator. No hardware is required today — remind participants the kit ships after they complete both surveys + a draft lesson. Pre-Survey: Open Pre-Survey. Shared Doc: Open Group Workspace.
8:40
Listen Contextualize 10 min
What Is Edge/IoT? The Two-Role Pattern
Brief overview: Edge vs. Cloud. Frame IoT as a conversation between two roles: sensor nodes (read + broadcast) and aggregator nodes (listen + aggregate + decide). 60-second sim demo — one simulator sends a radio number, the second simulator receives it and shows the value. Keep it short — participants want to build.
8:50
Do Reframe 15 min
Physical Computing Lives in the Browser Too
Poll: "To teach IoT I need…" (class set / 1:1 cart / maker lab / any browser). Breakout: #1 barrier to IoT in your building and what a simulator-first approach solves vs. loses. Groups post the trade-off in the Shared Workspace.
9:05
Break 10 min
Break #1
Stretch. Confirm everyone's MakeCode simulator is running and the LLM is open in a second tab.
9:15
Do Assemble Part A 25 min
Build a Sensor Node (Role A)
I Do (5 min): Live-code Role A in MakeCode — radio.setGroup(7), forever: radio.sendNumber(input.temperature()) + 2s pause. Point out the second simulator that appears. You Do (20 min): Everyone builds their own Sensor Node, picking a sensor (temp / light / accel / sound / compass). LLM-assist the code. Save the .hex — it flashes when the kit arrives. Engineering moment: idealized simulator temp vs. real CPU-die bias (preview for Fortify).
9:40
Do Assemble Part B 40 min
Paired: Aggregator Node (Role B) — Make the Halves Talk
Pair participants. One partner stays on their Sensor Node; the other builds the Aggregator — on radio.receivedNumber → running average / min / max / threshold alert → show number. Both must use the same radio group. LLM-assist the aggregation logic. Swap roles so everyone builds both halves. Everyone saves TWO .hex files. Post code + LLM prompts + fixes in the Shared Doc's Code Gallery.
10:20
Break 10 min
Break #2
Stretch. Facilitator checks group progress.
10:30
Do Fortify 25 min
Sensor + Aggregator Verification via CtM
Apply Check the Machine to the simulated pipeline: drag sensor sliders to extremes and confirm the running average responds; break the radio group and watch the Aggregator fall silent; remove radio.setGroup and observe the failure mode. Groups post Task / Before / After / Takeaway in the Shared Doc. Preview the hardware surprise (CPU-die temp) for the day the kit arrives.
10:55
Do Transfer 25 min
Two-Node IoT Lesson Design
Everyone drafts a two-node lesson: what does the Sensor Node sense? What does the Aggregator Node decide/display? Which students hold which role? NGSS standard + student CtM prompt. Add to the Shared IoT Lesson Idea Board. Pairs share and give feedback. "You just built a two-node IoT system in a browser, with AI as your co-pilot, and verified it like an engineer. Your students can do this too — even in schools with zero hardware budget."
11:20
Admin 20 min
Debrief + Resources + Post-Survey + Kit Unlock + Close
Kit unlock: confirm both surveys are complete and that a draft two-node lesson is posted in the Shared Doc — kits ship this week. Debrief wall in Shared Doc. Post-Survey: Open Post-Survey. Closing: series recap, what's next, stay connected.
Appendix: Master Resource List

Appendix: Master Resource List

Complete inventory of all links, files, and resources across the series.

Session Links (share in Zoom chat)

Workshop 1: AI for STEM (Apr 11)

Workshop 2: Verifying AI (Apr 18)

Workshop 3: Edge/IoT (Apr 25)

Website Resources (cxedhub.com)

Series and Workshop Pages

Participant Planned Agendas

Slides

Activity Resources

Admin and Setup

⚠️ Pre-Launch Checklist

⚠️ Pre-Launch Checklist

Verify before Workshop 1 (April 11):

✅ DONE

❓ VERIFY DEPLOYMENT (cxedhub.com)

❌ NOT YET CREATED

CRAFT PD Series · UCF DRACO Lab & UCF School of Teacher Education · Dr. Mike Borowczak & Dr. Andrea Borowczak