Teacher Daily Cards · Week 4
Can we prove our design works? — one page per session, generated from the current curriculum.
How to read these cards. Scan the header icons:
🔧 hardware-dependent ·
★ Gate/Milestone. In the run of show, each teacher cue is its own line;
Say: /
Ask: mark exact moves.
▽ below / ◆ at / △ above are the three differentiation tracks, and
⏱ If behind today tells you what to cut when the clock is short.
When it goes sideways, work down this ladder. You do not have to invent a fix mid-period. 1. The day's ⏱ If behind today line on the card — it names exactly what to cut, and on gate days what must NOT be cut. 2. The week's If it goes sideways card at the back of this pack, when the problem is bigger than the clock. 3. The Boost Card on the table, when it is one student stuck rather than the room. 4. Message Liz after class, not during. If your situation is not named anywhere, every recovery move here follows one principle: protect the thing being assessed, and let everything else move.
Week 4 · Day 16 · Gate/Milestone ★ 🔧IK Solve Card · Part B (arm run)
IK Solve Card, Part B — run your motion on the arm and measure the error at both ends.
Competencies this lesson builds
- ◆ PATTERN (Notice what repeats · Computational Thinking) — Gate/Milestone 3 — the same IK recipe runs at both endpoints. One method, two solves.
- ◆ REPRESENT (Convert between forms · Computational Thinking) — Paper angles become arm inputs (and the elbow offset: elbow = θ₂ + 90, using the SIGNED θ₂ — elbow-up means θ₂ = −θ₂′).
- ◆ CHOOSE (Pick the right tool · Computational Thinking) — The tool chain — Pythagoras, Law of Cosines, arctan — is selected and defended.
- ◆ DEBUG (Find & fix errors · Computational Thinking) — Students name where the error came from: math, measurement, arm calibration, or constraint.
- ○ MODEL (Name the quantities · Mathematical Modeling) — The measured landing is compared to the predicted one at BOTH ends.
- ★ JUSTIFY (Explain with evidence · Problem Solving) — The elbow-up choice and the error explanation are argued with evidence.
Warm-up · Daily Driver (on the board): To drive the arm to a point, do you type angles or enter (x, y) and press GO? (GO / angles)
Note: Plan ahead for Day 19: this week's rehearsal needs a fresh, non-expert audience — a neighboring class, a 9th grader, or a parent. Send the invite email today (it's Monday of Week 4) so someone is actually in the room Thursday.
Main work block · run of show (45 min) — one cue per line
- 0–5 · Entry Task
- GO.
- Enter (x, y), press GO — the app solves IK and handles the joint offset.
- Set up the arm stations.
- 5–12 · Frame Part B · the GO button
- They solved the poses last class; today they run them.
- Test the angle pair they wrote down BEFORE the calculation, typed into the manual joint fields with the conversion applied — elbow field = θ₂ + 90, and θ₂ is the SIGNED angle (elbow-up ⇒ −θ₂′), so (10, 8) takes −54.4.
- Entering a target and pressing GO runs the app's automatic IK solver instead, so a GO run that lands checks the target and the setup but cannot verify a hand solve that was wrong.
- Using the unsigned θ₂′ is the #1 manual-entry failure: raw θ₂′ lands the arm 43 cm off in this example, and θ₂′ + 90 is the elbow-down value, which the field accepts because it is inside the elbow's travel — the arm runs it and puts the end-effector on the far side of the base.
- Neither mistake gets caught for them.
- 12–36 · Run + measure
- Groups rotate to the arm, GO to each endpoint, run the full back-and-forth, and measure the miss at each end (cm).
- Compare against their Day-14 accuracy criterion — pass/fail, not just a number.
- 36–39 · Where did the error come from?
- Before they submit, a 90-second table talk on each end's gap: name the biggest source — the MATH (a wrong angle), the MEASUREMENT (reading the ruler / setting the target), the ARM (calibration, backlash, a loose joint), or the CONSTRAINT (the point was set wrong).
- Naming it now makes Day 17's one-change iteration deliberate, not guesswork.
- 39–45 · Submit + close
- Submit the completed IK Solve Card (Part A math + Part B measured).
- Spend any spare minute checking every student actually has a measured number, not a blank.
- Tags against PATTERN, REPRESENT, CHOOSE, DEBUG, MODEL, JUSTIFY (pass = all 6 checklist must-haves).
- Preview Day 17: “iterate — run, measure, adjust, run again.”
Students produce → IK Solve Card (Gate/Milestone — pass 6/6 must-haves)
Three tracks — below / at / above- ▽ Below — GO for both endpoints — enter each (x, y), press GO, measure the landing, name the error source. The offset is handled for them, so the card is the measurement and the reasoning.
- ◆ At — Hand-enter one endpoint (shoulder = θ₁, elbow = θ₂ + 90 with the SIGNED θ₂, so −θ₂′ for elbow-up) and GO the other, then compare the two landings — that is the offset felt rather than recited.
- △ Above — After both runs, straighten the elbow 2° and predict which way the tip moves BEFORE re-running. Only the elbow changes reach, and this is the head start on Day 17.
⏱ If behind today (35 min, or yesterday ran over): Setup eats the time here, not queueing — every crew arm and the front arm needs homing and calibrating BEFORE the bell, or you lose the first ten minutes to it. With four students per arm the runs themselves take about ten to twelve minutes. If you are short, ONE clean measured run at each end plus the error source is the must-do; push a precision or second run to Day 17.
Students will ask · what to sayAsk · “The arm missed by a lot.”
Say: First check you used GO, not hand-typed angles missing the +90 elbow offset. Then name the source — math, measurement, arm, or constraint. A clean miss you can explain beats a lucky hit.
Ask · “Do I re-do all my math?”
Say: No — carry your Part A poses onto the card, then add the measured accuracy and the error source. The complete card is what gets tagged.
Watch for- Hand-typing angles and forgetting the +90 elbow offset — push GO.
- Reporting 'close' instead of a measured cm value at each end.
Boost Cards: BC6, BC7, EC1, EC2 · arm setup (BC7)
Substitute today? Run it as a paper day — warm-up, work block, and exit ticket exactly as written above.
No arm needed: defer the Arm Check to the next arm day, or have students measure from the reference video (same motion, ruler in frame) — not yet filmed, so until it is, demo the motion by hand with a ruler or use the Day-11 aimer in the app, which simulates the same landing. The prediction and the error reasoning are unchanged — only where the measured number comes from.
Week 4 · Day 17 🔧Iterate: diagnose the miss, prove the fix
Measure the miss, work out what caused it, fix that — then prove it with a second run.
Competencies this lesson builds
- ○ TEST (Prediction vs. reality · Mathematical Modeling) — Run 1 is a real measurement against the Day-13 criterion — it passes or it doesn't.
- ◆ DEBUG (Find & fix errors · Computational Thinking) — Students locate the error before changing anything. Guessing at fixes is the failure mode.
- ○ REFINE (Improve based on data · Mathematical Modeling) — Change exactly ONE thing, driven by the data — not three things at once.
- ★ ITERATE (Improve from feedback · Problem Solving) — Run 1 → adjust → run 2. The improvement is the evidence, and it goes in the final deck.
Warm-up · Daily Driver (on the board): Your criterion is “within 1.5 cm.” Run 1 missed by 2.3 cm. Did it pass? (yes/no)
Note: The arms run a correct solve faithfully, so a small miss is not worth a joint change and the day should not send students there first. Deliberately no accuracy figure anywhere in this day — it would need re-measuring per fleet, and the student's own criterion is the honest yardstick. The old version opened on “your first run almost never lands perfectly” and went straight to nudging, which cures almost nothing they will actually meet.
Main work block · run of show (45 min) — one cue per line
- 0–5 · Entry Task
- No — 2.3 > 1.5, so it misses the criterion.
- That is what a number buys you: the answer is not a matter of opinion.
- 5–16 · Teacher-led · the miss has a size, and the size has a name
- Say the headline first: the arm runs your solve faithfully, so the question is never how far off am I on its own — it is how far off, does it repeat, and does it clear the criterion I set.
- Put the table up: small and clears the criterion means you pass, leave it; small and fails it is setup; ≈ 20 cm is the IN/CM toggle; ≈ 43 cm is θ₂′ typed raw into the elbow field; different-every-run is slack.
- Then the rule that comes before any fix — run it AGAIN before changing anything, because a miss that does not repeat cannot be cured by a number.
- 16–36 · Diagnose and re-run on the arm
- Groups measure run 1, re-run untouched to test repeatability, name the cause, fix THAT, and re-measure.
- Most will find units, setup, or hand-entry.
- A crew that has genuinely ruled all of it out and still has a repeatable offset may bias the elbow — but do not hand them a cm-per-degree figure.
- Use each group’s own arm and target.
- Start with an unchanged repeat and the student’s criterion.
- If both runs pass, keep the command and document that result.
- For a confirmed repeatable radial miss that remains after the approved calibration procedure, use Activity 5’s own-arm record.
- Record manual J₁ and J₂ and actual x and y before and after a teacher-checked 1° elbow test, with the shoulder fixed.
- Students calculate reach as √(x² + y²), then signed reach change per degree.
- The rate varies with pose and can be very small near a straight arm.
- If the change is unclear, repeat the measurement and investigate before dividing or scaling.
- If the test passes, stop.
- Otherwise estimate additional degrees from the remaining reach difference after the test, so the first degree is not counted twice.
- Agree on a small next step within the installed limits and a checked path.
- Judge the full distance to the target after every trial because elbow motion also changes tip direction.
- Check the measured record, calculation, explanation and retest, not a common adjustment angle.
- If the arm is unavailable, leave measurement pending and arrange a retry.
- Do not substitute another group’s rate.
- 36–45 · Debrief + preview
- Collect causes on the board, not error sizes.
- The number nobody can present is “we changed something and it got better.” Land the joint rule while you are there: the ELBOW sets how far the tip sits from the base, and the SHOULDER rotates that reach toward the target direction.
- A miss that is simply too short is radial, so it belongs to the elbow.
- Preview Day 18: the deck, and this story is part of it.
Students produce → Your run misses by 20 cm, the same amount every time. What is the first move?
Three tracks — below / at / above- ▽ Below — Give them the diagnostic table and one scenario at a time. Repeat the run before naming a cause, then walk the checks: the math, the units, the fields, the setup, and the hardware. The 20 cm and 43 cm figures are clues from those particular examples, not codes to read off any miss. They classify rather than compute.
- ◆ At — Measure, run again to test whether it repeats, name the cause, fix it, re-measure — and say what the result told them.
- △ Above — Predict the SIZE of the improvement before running, then explain why the shoulder cannot change reach at all while the elbow can — and whether biasing a command is a fix or a workaround compared with RECALIBRATE.
⏱ If behind today (35 min, or yesterday ran over): The must-do is measure → does it repeat → what caused it → fix and re-run. If the arm queue is long, each group documents one diagnosis and one re-run rather than exploring. What cannot be cut is the second run: without it there is no iteration to present on Day 18.
Students will ask · what to sayAsk · “It got worse after I changed it.”
Say: Good data — that tells you the direction. But first: did you establish the miss repeated before you changed anything? If not, you may have been chasing slack.
Ask · “We passed our criterion on run 1. Is there nothing to iterate on?”
Say: That IS the result, and it is presentable. Report the measurement against the criterion, say what you checked to be confident it was real, and use the second run to show it repeats. A run that confirms you were right is still evidence — and changing a passing result to look busy is the one move that loses marks.
Ask · “Can we just nudge the elbow until it lands?”
Say: Only after you have ruled out units, offset and setup, and only if the miss repeats. Otherwise you are hiding a cause rather than fixing it — and the same arm will miss differently tomorrow.
Watch for- Nudging a joint before establishing the miss repeats. That is chasing noise.
- Changing several things at once. One variable per run.
- Reporting “close” instead of a measured cm value.
Boost Cards: BC8, BC6, BC7 · arm setup (BC7)
Substitute today? Run it as a paper day — warm-up, work block, and exit ticket exactly as written above.
No arm needed: defer the Arm Check to the next arm day, or have students measure from the reference video (same motion, ruler in frame) — not yet filmed, so until it is, demo the motion by hand with a ruler or use the Day-11 aimer in the app, which simulates the same landing. The prediction and the error reasoning are unchanged — only where the measured number comes from.
Week 4 · Day 18Build the deck
Build a 7-slide deck that lands with your user in the room.
Competencies this lesson builds
- ★ JUSTIFY (Explain with evidence · Problem Solving) — Every claim in the deck is backed by a number the crew actually measured.
- ★ CREATE (Manage constraints · Problem Solving) — Seven slides is a constraint. Deciding what to cut is the work.
Warm-up · Daily Driver (on the board): A slide says “the arm was pretty accurate.” One thing turns that from an opinion into evidence. What is it? (one word)
Note: Students run this one themselves. The app carries the audience framing and the 7-row template, each row naming the artifact the slide comes from, so there is no front-of-room block to deliver. Your job is circulating, not presenting — and the three fields at the bottom of the day give you a read on every deck before Day 19 rather than at the gate.
Main work block · run of show (45 min) — one cue per line
- 0–5 · Entry Task
- A number — one they actually measured, with the criterion they judged it against.
- That is the standard every claim in the deck has to meet.
- 5–38 · Independent build
- Students work the day themselves: the audience note, the 7-row template, the opening-slide question, then build.
- Every row of the template names where the content already is — Brief, IK Solve Card, Part B, the two calibration runs — so this is assembly, not writing.
- Circulate.
- The one intervention worth making is killing any slide that opens with “inverse kinematics uses…”, and the app's opening question sets that up for you.
- 38–45 · Hand in + exit
- They paste the deck link, write the ONE thing their user needs to understand, and name their weakest slide.
- Read those three fields tonight — that is your Day 19 grouping, and it is the earliest warning you get if a deck does not exist.
Students produce → When the user is your audience, the best first slide…
Three tracks — below / at / above- ▽ Below — Work the template in the app one row at a time — each row names the artifact the slide comes from, so it is an assembly job, not a writing job. They fill in their own numbers.
- ◆ At — Build all seven, then answer the three hand-in fields honestly — especially which slide is weakest.
- △ Above — Draft the deck, then cut it. Which slide could go entirely without the user losing the thread? Defend keeping the one you nearly cut.
⏱ If behind today (35 min, or yesterday ran over): The must-do is a complete rough deck and the three hand-in fields. A rough deck that exists beats a polished one that is half-built — Day 19 is where it gets better, and it cannot rehearse what is not there.
Students will ask · what to sayAsk · “Where do I build it?”
Say: Whatever slide tool the class uses. The seven rows are the structure; the tool does not matter, and the link goes in the hand-in field so nothing gets lost.
Ask · “Do I need a slide for the interview and the sketch?”
Say: No — they are folded in. The interview line goes on slide 1 and the sketch on slide 2. The Day 20 rubric scores both, which is why the template puts them somewhere rather than leaving them out.
Ask · “Mine is rough.”
Say: Good. Rough and complete is exactly what Day 19 needs — rehearsal cannot fix a slide that is not there.
Watch for- A deck that opens with the method rather than the user. It is the single most common miss, and the fix is one sentence.
- Claims with no number behind them. Every slide that says it worked has to show what was measured.
- Students who have not pasted a link by the end. If it is not in the field, nobody can look at it before the gate.
Boost Cards: BC7, BC6, EC3, EC1
Substitute today? Run it as a paper day — warm-up, work block, and exit ticket exactly as written above.
There is no arm today, so nothing to defer.
Week 4 · Day 19 🔧Peer review & rehearsal
Present to a partner, mark up the rubric, and rehearse the live demo.
Competencies this lesson builds
- ★ ITERATE (Improve from feedback · Problem Solving) — Peer feedback against the rubric, then revise — the rehearsal exists to change the deck.
- ★ JUSTIFY (Explain with evidence · Problem Solving) — Students defend their solve out loud and answer the question they didn't expect.
Warm-up · Daily Driver (on the board): True/false: the best opening slide for your user leads with what the arm does for them.
Note: Line up the outside listener in advance — email a neighboring teacher (or invite a parent / a 9th-grade class) on the Monday of Week 4 so a fresh, non-expert audience is here for today’s rehearsal.
Main work block · run of show (45 min) — one cue per line
- 0–5 · Entry Task
- 5–20 · Trade presentations
- Present to a partner AND to someone who does NOT know what IK is (another class, a 9th grader, a parent).
- If they get WHY it matters, you truly understand it.
- Both mark the rubric.
- 20–40 · Rehearse the live demo
- Practice the live arm run — the riskiest, unfakeable part.
- GO to each endpoint; catch what breaks now: pairing, the +90 offset if anyone hand-enters angles, the sweep.
- Have a backup measurement ready.
- 40–45 · Exit + preview
- Rehearsal surfaces problems before the audience does.
- Preview Day 20: “the real thing.”
Students produce → The point of rehearsal is to…
Three tracks — below / at / above- ▽ Below — Rehearse to one partner with presenter notes in hand, then run the arm once with a crewmate reading the endpoints aloud.
- ◆ At — Present to a partner AND a non-expert listener, both marking the rubric, then rehearse the live run.
- △ Above — Have the listener ask one unscripted question and answer it from their own numbers — the question they most fear is the one to rehearse.
⏱ If behind today (35 min, or yesterday ran over): The live demo rehearsal is the must-do: it is the only unfakeable part of Day 20 and the only one that fails in front of an audience. Cut the partner trade to a single pass with the rubric, and if the outside listener did not materialize, borrow a student from another table who has not seen the deck. If the arm queue is long, each group rehearses the run once and takes the backup photo or video right then — that is what keeps a Day-20 hardware failure from taking the story with it.
Students will ask · what to sayAsk · “What if the arm fails on presentation day?”
Say: That’s why we rehearse. Have your measured results and a photo/video ready so the story survives a hardware hiccup.
Watch for- Skipping the live-demo rehearsal because ‘the slides are done.’ The demo is the risk.
Boost Cards: BC7, BC6, EC3, EC1 · arm setup (BC7)
Substitute today? Run it as a paper day — warm-up, work block, and exit ticket exactly as written above.
No arm needed: defer the Arm Check to the next arm day, or have students measure from the reference video (same motion, ruler in frame) — not yet filmed, so until it is, demo the motion by hand with a ruler or use the Day-11 aimer in the app, which simulates the same landing. The prediction and the error reasoning are unchanged — only where the measured number comes from.
Week 4 · Day 20 · Gate/Milestone ★ 🔧Final Presentation
Milestone 4 — present to your user and run the arm live.
Competencies this lesson builds
- ★ DESIGN (Create for user needs · Problem Solving) — Gate/Milestone 4 — the solution is presented to the user it was built for.
- ★ CREATE (Manage constraints · Problem Solving) — The live demo runs under real constraints, in front of a real person.
- ★ JUSTIFY (Explain with evidence · Problem Solving) — The math is defended with measured evidence.
- ○ REFINE (Improve based on data · Mathematical Modeling) — The calibration story shows the solution improving from data.
- ★ ITERATE (Improve from feedback · Problem Solving) — Run 1 → adjust → run 2 is shown, not just claimed.
- ○ TEST (Prediction vs. reality · Mathematical Modeling) — The arm runs live and the landing is measured in the room.
Warm-up · Daily Driver (on the board): In one word: what proves your solution works — a slide, or the live arm demo?
Note: Two-part day across the week: the deck and My Work record come together on Days 18–19. Today is the live presentation, the arm demo, and rubric scoring — not deck-building.
Main work block · run of show (45 min) — one cue per line
- 0–5 · Entry Task + setup
- demo.
- Set up the arm; groups ready their My Work record (Brief, FK Card, IK Card, calibration runs).
- 5–40 · Presentations (simultaneous small-group)
- Groups of 4–5 present to each other at once while you circulate; each student gets 5–7 min.
- Run the arm live through the full motion.
- Score on the 8-section rubric (W4D5 exemplars for calibration).
- 40–45 · Whole-class share + close
- Call 2–3 groups to present to the whole class as a capstone.
- Collect final reflections.
- Score the 8-section rubric live as each group presents — a performance cannot be scored after it is over.
- Tags against DESIGN, CREATE, JUSTIFY, REFINE, ITERATE, TEST.
Students produce → Final Presentation (Gate/Milestone — pass rubric)
Three tracks — below / at / above- ▽ Below — Presenter notes in hand — one sentence per slide, written on Day 19 — and a crewmate cued to drive the arm so they can talk through the run instead of operating it.
- ◆ At — The 7 slides plus the live run, presented and scored on the 8-section rubric.
- △ Above — Take an unscripted question from the audience and answer it with their own numbers — the JUSTIFY move the rubric's top band asks for.
⏱ If behind today (35 min, or yesterday ran over): Everyone presents and everyone gets scored — that is the gate, and neither half is optional. Cut the whole-class capstone share-out first; the simultaneous small-group round is what covers the rubric. If the clock is genuinely gone, hold each presentation to 4 minutes on a hard timer and score from the small-group round only. Protect the live run for every group that still has one — it is the unfakeable part. If the arm is down, measured results plus the Day-19 rehearsal photo or video carry the demo; that path is on the week's Contingencies card.
Students will ask · what to sayAsk · “Do I have to present to the whole class?”
Say: Most present in small groups simultaneously; a few are selected for the whole-class share-out. Everyone presents.
Watch for- Presentations that recite the math but never say why it matters to the user. That’s the difference between a 3 and a 5.
Boost Cards: BC7, BC6, EC3, EC1 · arm setup (BC7)
Substitute today? Run it as a paper day — warm-up, work block, and exit ticket exactly as written above.
No arm needed: defer the Arm Check to the next arm day, or have students measure from the reference video (same motion, ruler in frame) — not yet filmed, so until it is, demo the motion by hand with a ruler or use the Day-11 aimer in the app, which simulates the same landing. The prediction and the error reasoning are unchanged — only where the measured number comes from.
Week 4 · ContingenciesIf it goes sideways
Three recovery moves for this week — preserve rigor, ease the crunch. Reach for this card when the problem is bigger than the clock; if it is only the clock, the day's ⏱ If behind today line is the faster fix.
- The arm fails on presentation day — Measured results + a photo or video of a rehearsal run carry the demo. The rubric scores the plan and the math — a hardware hiccup doesn't sink the grade.
- Half aren't deck-ready by Day 19 — The 7 slides map to 8 sections they already have in My Work (Brief, FK, IK, calibration). Assemble from those — an honest rough deck beats a polished late one.
- The outside audience falls through — Present to the class as the outside audience, or record it for one. The point is a non-expert hearing WHY it matters.
- The app or the internet is down — The arm is on its own wi-fi network, so the Day 16 run and the Day 17 calibration go ahead. Hand out the printed IK Solve Card on Day 16 and the Final Presentation template on Day 20 and tag the paper. Days 18 and 19 move to paper storyboards.
Generated from the live curriculum · npm run gen:printables