LGR22 and Slöjd: AI micro-tools that protect making

Four teacher-led routines for planning, language, evidence and safety

A Slöjd teacher guiding pupils while using an AI planning template on a laptop

Why Slöjd tests AI

Slöjd is a hard test for AI because the subject is deliberately resistant to shortcuts. The learning is embodied: how a pupil holds a saw, adjusts seam allowance, chooses grain direction, or changes a plan when the material behaves unexpectedly. That is precisely why it is a good place to use AI carefully. If AI can support planning, language and reflection without stealing the making, it can support almost any subject responsibly.

In practice, the risk in Slöjd is not that pupils will “cheat” by generating a finished product; it is that adults may over-trust confident text and diagrams. A neat-looking pattern, a plausible risk list, or a polished reflection can mask unsafe or unrealistic advice. If your school is already tightening routines for transparency and compliance, it helps to connect Slöjd work to wider guidance such as this explainer on EU AI Act and LGR22 compliance, because the principles (human oversight, safety, documentation) fit Slöjd naturally.

LGR22: what to protect

Under LGR22 in Slöjd, the essentials worth protecting are not “outputs”, but the process and judgement behind them. The hantverksprocessen matters because pupils learn to plan, test, evaluate and iterate. Materialkunskap matters because pupils learn that textiles, wood and metal have properties that shape design choices. Safety matters because tools, machines, heat and dust are real hazards. Reflection matters because pupils must explain choices, assess quality and consider improvements.

AI can support these aims when it is used as a prompt for thinking rather than a replacement for thinking. A useful rule is: AI can help you prepare the runway, but pupils must fly the plane. If you are building inspection-ready routines across subjects, you may find it helpful to align language and documentation with your wider LGR22 approach, as outlined in LGR22 Section 2 throughlines and micro-tools.

Non-negotiable red lines

AI must not become an unaccountable “silent instructor” in the workshop. In Slöjd, the red lines are simple and strict.

AI must not give step-by-step instructions for machine operation that bypass your local routines, demonstrations, or permission systems. It must not be used to overrule safe practice, tool limits, or supervision ratios. It must not generate “authoritative” risk assessments without teacher verification and adaptation to your specific room, tools and pupils. It must not be used to fabricate process documentation, reflections, or photographs; if the evidence is not real, it undermines assessment and safety culture.

Finally, AI must not be treated as a source of truth about materials, allergens, finishes, or tool compatibility. Use it to draft questions and checklists, then verify with trusted sources and your own expertise.

Micro-tool 1: Making lesson planner

A well-designed AI lesson planner can protect the making by shifting your time from writing worksheets to coaching decisions. The key is to plan for process goals at E/C/A levels and to embed “stop points” where pupils must show their thinking before they continue.

Example: Zip pencil case, Åk 6, 80 minutes. You might ask AI to draft a lesson flow with three process checkpoints: planning, construction, evaluation. The output you want is not a script, but a flexible structure: a short demonstration, a guided planning moment, independent making, then a reflection routine.

At E level, process objectives might focus on following a simple plan, using basic tool handling safely, and explaining one choice (“I chose this fabric because…”). At C level, objectives can emphasise adapting a plan when problems appear (zip alignment, seam allowance, fabric fraying) and describing how changes improved function. At A level, objectives might include evaluating quality against criteria, comparing alternative methods (boxed corners vs flat base), and giving a reasoned improvement plan.

Keep the AI honest by asking it to produce “teacher prompts” rather than pupil instructions. For instance, instead of “Sew the zip like this”, you want “Ask pupils to show their pinning plan before stitching; check zip pull direction; confirm seam allowance”. This approach also matches the broader idea of time-saving, teacher-in-the-loop workflows described in LGR22 three years on: gap-to-tool map.

Micro-tool 2: Pattern converter SV↔EN

Slöjd classrooms often use mixed resources: Swedish instructions, English YouTube terms, international patterns, and pupils who switch languages mid-sentence. A small AI “converter” can translate knitting and sewing notation both ways, while preserving meaning and flagging ambiguity.

The trick is to treat conversion as a bilingual literacy routine, not a one-click answer. Ask AI to produce a three-column table: original line, converted line, plus a “check me” column for terms that vary (for example, US vs UK crochet terms; abbreviations that differ by pattern author). Then you run a quick classroom glossary routine: pupils add two new terms to a shared wall or booklet, with a sketch or photo of the technique.

This pairs well with retrieval-friendly displays; you can connect it to the broader idea of vocabulary and dual coding from AI inclusive classroom displays by keeping the glossary visual and practical. The goal is not perfect translation, but confident communication about techniques and materials.

Micro-tool 3: Documentation prompts

Process documentation is often where Slöjd assessment becomes manageable—and where AI can help without touching the product. Instead of asking pupils to “write a log”, use AI to generate short, timed prompts that capture decisions at the moments they happen.

A simple structure is: planning log, decision points, photo captions, reflection questions. In the zip pencil case lesson, pupils might answer two planning prompts before they collect materials (“What will success look like for you today?” and “Which step do you think will be hardest, and what will you do if it goes wrong?”). During making, you pause the room twice for a 60-second decision capture (“What did you change from your plan?” and “What did you measure or test before sewing/cutting?”). At the end, pupils add one annotated photo and respond to two reflection questions that link to criteria (“Which part shows the best quality, and why?” and “What would you improve next time?”).

AI’s role is to draft prompts at different language levels, including sentence starters, and to keep them aligned to process goals. Pupils must supply the content, photos and thinking. If you want a wider set of safe classroom routines for AI use, the micro-routine approach in INSET day AI workshop: three micro-routines adapts well to Slöjd because it emphasises repeatable habits.

Ready to Revolutionise Your Teaching Experience?

Discover the power of Automated Education by joining out community of educators who are reclaiming their time whilst enriching their classrooms. With our intuitive platform, you can automate administrative tasks, personalise student learning, and engage with your class like never before.

Don’t let administrative tasks overshadow your passion for teaching. Sign up today and transform your educational environment with Automated Education.

🎓 Register for FREE!

Micro-tool 4: Safety-first risk workflow

Risk assessment is where Slöjd needs the most caution with AI. Used properly, AI can help you avoid blind spots by generating a structured checklist; used poorly, it can create a false sense of security. The safest workflow is: AI drafts, teacher verifies, room-specific edits, then a pupil-facing safety briefing.

For textiles, you might prompt AI to list hazards such as needle injuries, rotary cutter use, hot irons, trailing leads, allergies to fibres, and posture fatigue. For wood, it should include splinters, dust, sharp edges, clamping failures, and tool storage. For metal, it should include sharp swarf, heat, eye protection, and noise. You then add your mitigations: supervision points, PPE requirements, machine permission steps, first-aid location, and the exact room set-up (walkways, “hot zone” marking, extraction use, and where bags and coats go).

Ask AI to output four columns: hazard, who might be harmed, mitigation, and “teacher check”. That last column forces you to confirm what is actually true in your space. If you maintain a school-wide safety protocol for AI, align it with your broader approach to safe roll-out, such as the routines described in the ECT/NQT AI operating manual, because consistency matters when staff rotate rooms.

Quality control checks

Before you use any AI output in Slöjd, run quick checks for accuracy, feasibility, inclusivity and sustainability claims. Accuracy means checking measurements, tool names and technique steps against trusted references and your own practice. Feasibility means asking, “Can this be done in our time, with our equipment, and with this class?” Inclusivity means ensuring alternatives exist for pupils with motor difficulties, sensory needs, or language barriers, without lowering the process challenge. Sustainability claims need special care: if AI says a material is “eco-friendly”, require evidence, and prefer local purchasing guidance and known certifications.

A practical habit is to add a one-line footer to any AI-generated teacher document: “Drafted with AI, verified and adapted by the teacher for this room and class.” It keeps your practice transparent and professional.

30-minute department set-up

You can set this up in half an hour if you keep it simple. Create four templates (planner, converter, documentation prompts, risk workflow) and store them in a shared drive with clear version dates. Build a small prompt library so staff do not start from scratch, and include a “do not use AI for…” note at the top of each template. Decide where pupil documentation lives (paper booklets, a digital portfolio, or both) and how photos are handled safely. Finally, write a short transparency note to pupils explaining what AI is used for in Slöjd: planning support, language support, and reflection prompts—not doing the making, not writing reflections for them, and never replacing safety instruction.

If you already review acceptable use annually, fold these Slöjd-specific rules into your wider process, such as the AI acceptable use policy refresh checklist.

Appendix: prompt templates

Use these as copy-and-adapt starters. Replace bracketed text with your context, and keep outputs short.

For the lesson planner: “You are helping a Slöjd teacher plan an 80-minute making lesson for Åk 6: [project]. Create a lesson flow with three process checkpoints and E/C/A process objectives focused on planning, safe technique, and evaluation. Do not give machine-operation instructions. Include teacher questions that prompt pupil decision-making.”

For the pattern converter: “Convert the following knitting/sewing instructions from Swedish to English (UK terms) while preserving abbreviations where possible. Output a table: Original | Converted | Ambiguities to check. Flag any term that could differ by region or pattern author. Text: [paste].”

For documentation prompts: “Create pupil-friendly process documentation prompts for [project], suitable for mixed language levels. Provide: two planning prompts, two mid-process decision prompts, one photo caption prompt, and two reflection questions linked to quality and improvement. Include sentence starters and keep each prompt under 20 words.”

For the risk workflow: “Draft a risk assessment checklist for a Slöjd lesson using [textile/wood/metal] tools: [list tools]. Output columns: Hazard | Who may be harmed | Mitigation | Teacher check (room-specific). Avoid giving operational instructions for machines; focus on hazards, supervision points, PPE, and room set-up prompts.”

May your workshops stay calm, creative, and safely hands-on.
The Automated Education Team

Table of Contents

Categories

Teaching

Tags

Lgr22 Slojd Ai tools

Latest

Alternative Languages