AI PBL Bootcamp · Free · Three Mondays from 26 October
Design

Build the project calendar and milestone map

An extended project lacks a realistic sequence from launch through inquiry, drafting, critique, revision and public culmination.

AI does

Schedule drafting, dependency checks and alternative scenarios.

You do

Judge local pacing, workload, transitions and where direct instruction or flexibility is needed.

1

Research

Why this matters, and what good looks like

A milestone is a point where students show work for a check. A dependency is something that must happen before another task can start. A useful calendar connects these with the time actually available. The sources below are professional planning guidance and a practitioner framework, rather than studies of this scheduling method. They establish no AI benefit or preferred time allocation. Protecting critique and revision is a design choice in this workflow. Local pacing remains your decision.

  • The Project Planner brings learning goals, launch decisions, products and daily lessons into planning. Include teaching alongside production.
  • PBLWorks teaching practices involve teachers and students in schedules, tasks and checkpoints. Check the proposed sequence with those doing the work.
  • The HQPBL framework describes extended inquiry and sharing during development and at the conclusion. Keep development visible before the final event.

Where the evidence comes from

Strong for PBL project management; AI benefit unvalidated.

2

Workflow

Brief it, steer it, check it
  1. Establish the available time

    Use an approved D04 brief or equivalent teacher documents. Bound this pass to one short project, six meetings and three outcomes. Keep private learner records outside AI. If active planning exceeds 60 minutes, stop and reduce scope.

    Prompt 1 · Check inputs
    CONTEXT
    Approved project brief, including outcomes, product, sharing arrangements and required teaching: [brief].
    Teacher-verified available-time ledger, with version date, each meeting date, usable minutes after transitions, unavailable dates, fixed deadlines and source references: [ledger].
    Teacher estimates of task durations, dependencies, readiness checks, critique, revision and contingency needs; confirmed responsible roles and student choices: [requirements].
    Existing calendar to improve, or none: [existing calendar].
    REQUEST
    Inventory the inputs for a dated calendar. Separate teacher-verified availability from estimated task time. Identify contradictions, stale versions and missing decisions before proposing a schedule.
    QUALITY BAR
    Use only supplied documents. No student identities or private support facts. If the dated availability ledger, approval status or required teaching is missing, return HOLD and questions only. Treat unknown time and commitments as unavailable. Do not invent dates, audiences, role acceptance or readiness. A supplied verification is a teacher report, not your independent verification.
    FORMAT
    Return an input table: item, exact source reference, supplied fact or estimate, conflict or HOLD, teacher action. End with decisions needed before scheduling.
  2. Compare feasible sequences

    Before accepting any capacity PASS or shortfall HOLD, independently recalculate task sums, each date's usable minutes and separate reserve from supplied figures. Deduct transitions once. Record calculations before approving the corrected inventory. Another AI check or retry cannot replace yours.

    Prompt 2 · Sequence
    CONTEXT
    Exact step 1 inventory approved by the teacher: [approved inventory].
    Teacher corrections, selected source versions, confirmed ledger and independent arithmetic record showing task sum, each date's gross minutes minus transitions once, total usable time and separate contingency, completed outside AI before inventory approval: [decision addendum].
    REQUEST
    Propose two sequencing options, or explain why only one fits. For each, place launch, inquiry, instruction, drafting, critique, revision and culmination in the supplied dated slots. Show dependencies and readiness checks before dependent work.
    QUALITY BAR
    Stop with HOLD and questions only if approval, dated usable minutes, teaching needs, task estimates or the teacher's independent arithmetic record are absent. Your calculations or another AI retry cannot supply that record; a supplied record remains a teacher report, not your verification. Never schedule unavailable time or treat task estimates as timetable facts. Show used and remaining minutes per date, including transitions already deducted in the ledger. Reserve teacher-specified critique, revision and contingency separately. If demand exceeds capacity or a prerequisite is unresolved, report the shortfall and teacher choices to reduce scope or verify more time; do not silently compress learning. Mark all options PROPOSAL. Do not select one or reassign confirmed roles.
    FORMAT
    For each option give a dated table: milestone/task, estimated minutes, prior dependency, instruction/readiness check, responsible role, checkpoint, remaining minutes. Add totals, constraint conflicts and choices preserved.
  3. Check the selected calendar under pressure

    With colleagues and students, choose an option. Before approval, independently recompute each date's allocated work plus separate reserve against usable minutes; record totals, changes, confirmed roles and a disruption. Reduce scope or revise dates without removing critique or revision.

    Prompt 3 · Stress check
    CONTEXT
    Exact step 2 option selected and edited by the teacher: [selected calendar].
    Teacher decision addendum confirming selection, student/colleague input, current ledger, protected windows, one disruption to test and independent per-date work-plus-reserve calculations against usable minutes before calendar approval: [selection addendum].
    REQUEST
    Check the selected calendar and compare it with the specified disruption. Trace effects through dependencies, instruction, checkpoints and workload. Offer bounded contingency choices for the teacher.
    QUALITY BAR
    Return HOLD and questions only if selection, the current ledger or the teacher's independent per-date arithmetic record is missing. Another AI check cannot substitute for that record. Keep base schedule and hypothetical disruption separate. No invented spare periods, homework, bookings or role acceptance. Use reserved contingency first only as a proposal. Keep critique and revision protected. Flag instruction that would occur after the task needing it. Show unresolved readiness as a gate, never assume students are ready. If no feasible recovery exists, state HOLD and the shortfall. Do not adopt a recovery choice.
    FORMAT
    Return base-calendar issues, a disruption impact table with source references and minutes, then teacher choices and unresolved gates. Label proposed changes explicitly.
  4. Audit the teacher-selected result

    Resolve the stress check outside AI. Independently recalculate changed capacity and allocations, recording work, transitions deducted once and separate reserve before approving the final calendar. Verify dates with responsible people. Release only your corrected version after checking the audit yourself.

    Prompt 4 · Final audit
    CONTEXT
    Exact step 3 stress-check output reviewed by the teacher: [reviewed stress check].
    Teacher-selected final calendar and milestone map: [final calendar].
    Explicit teacher edits, recovery choice, source ledger and brief versions, role confirmations, unresolved decisions and independent recalculation of final per-date capacity, work and separate reserve before approval: [final decisions].
    REQUEST
    Audit the selected result against the supplied sources and stress-check issues. Check dates, capacity arithmetic, dependencies, required instruction, readiness gates, roles, student choices, checkpoints, protected critique/revision and contingency.
    QUALITY BAR
    If the teacher's independent final arithmetic record is absent, return HOLD and questions only; another AI calculation cannot replace it. HOLD missing, stale or unsupported claims. Never certify actual availability or student readiness. Distinguish source-matched teacher decisions from proposals and verified facts reported by the teacher. Do not rewrite the calendar or approve release. If estimates exceed usable minutes, critique/revision disappears or unavailable time is used, identify a blocker. Only the teacher can release the result after human checks.
    FORMAT
    Return blockers with exact calendar line, source evidence and teacher action; source-matched items with references; and remaining human checks. If none are found, say no blockers found in this comparison, without declaring readiness.

Check before you use it

  • Does each date and usable minute total match your current, teacher-verified availability ledger?
  • Have you independently recomputed task sums, per-date capacity and allocations, deducting transitions once and separating reserve before approval?
  • Does required instruction precede dependent work, with readiness checked rather than assumed?
  • Are critique, revision and contingency visible and protected through the selected disruption response?
  • Have responsible people confirmed roles and students retained their agreed choices?
  • Have you removed private learner information and resolved every HOLD before releasing your calendar?

Stop rule

Stop if the calendar removes critique/revision, assumes unavailable time or treats estimates as timetable facts.

3

Reuse it

So the next one takes minutes
Suitable

Build a skill

Save this as a reusable instruction you can run on any project.

Repeat the bounded input check, sequence comparison and audit with a fresh verified ledger. Teachers still select and release each calendar.

Not suitable

Build an agent

Chain the steps, with a checkpoint where you approve.

This bounded planning task needs local timetable and instructional decisions. Autonomous rescheduling would exceed its authority.

Help us improve this method

Anonymous, and it takes under a minute.

Did you get something usable?
How useful was this method?
1 = wasted my time · 5 = I'll use it again
Where did it break down?

Please don't include names, school names or anything about individual students.

No account, no email, nothing stored about you.

Use cases are starting points with one perspective. Disagree with the process, that's expected. Make it yours.

Co-funded by the European Union under Erasmus+ KA210-SCH. No student data. No teacher material stored.