Our approach: what we believe, and how we teach

Everything we do comes from a few beliefs. From them come our values, what students learn, how they learn, and what we do next. We look at this page again every year, and grow it step by step.

What we believe

Young people can make technology, not only use it

Start small, one real step at a time, and understanding and confidence grow. That is the name START A BIT.

Software, AI and hardware are one system

We teach the whole working system: real things that measure, decide and act, with AI inside where it helps. Why we build real things.

Build the machine. Keep the know-how.

AI may help to build, but the know-how stays the student's: they can explain, test and change what they made.

Everyone belongs

Inclusion is everyone's job. We explain, we do not show off, and quiet students have as much room as loud ones. Our agreement.

Growing independence, not promises

We do not promise careers. We promise students who make more of their own technical decisions, build working systems and check their own understanding.

What we value

Questions are welcome. Asking, and saying "I don't understand yet", is how we learn.
Plain words. We explain technical words, so everyone can follow.
The real thing decides. We test on the real board: what works, what fails, and why.
Honest understanding. We say what we understand and what we still need to check.
Safety and care. For people, tools, parts and the room.
Help without taking over. We share what we know, and leave the work in its owner's hands.

So, what students learn

Three learning modes

Three overlapping circles: Explore (Micro:bit, blocks, first circuits), Build (Arduino, Python and C++, sensors, motors) and Create (Raspberry Pi, cameras, AI, team machines); plan, build, test and fix in the middle, with arrows going round through all three

Explore (try and ask), Build (make things that work) and Create (develop your own ideas) overlap. They are not levels to pass: most projects mix them, and students come back to all three.

Five learning areas

Logical thinkingsteps, conditions, testing an idea
Programmingwrite, read and fix programs
Electronicsconnect parts and understand them
Systems thinkinghow software, AI and hardware affect each other
AI competenceuse AI with clear limits and check its answers

Students go deeper in each learning area every time they come back to it. Together they build mindset, habits, confidence and know-how, not only skills. The learning map shows them step by step.

So, how they learn

Small groups, at most eight students, taught in English.
One story at a time: each project grows in steps, each with "I can" goals students can check.
You first: students try, then may use AI in one or two ways the teacher agrees.
Real boards: Micro:bit (blocks, then Python), Arduino (C++), Raspberry Pi (Python, cameras, AI); ESP32 and Raspberry Pi Pico when students are ready.
Practice: quizzes, bug hunts and a code editor in the browser for Python, C/C++ and JavaScript.
Extra help from the teacher when a student is really stuck.
Show it without AI: at the end, students show and explain their project themselves.

What comes next

We grow with new projects (a digital pet, a windscreen wiper, a robot arm, a sumo robot, an AI car with a camera) and new boards when students are ready. Every year we look again at the boards, the projects and this page, with the students' work and the guides and studies below. The aim stays the same: students who make more decisions, build working systems and check their own understanding.

What we learn from

We keep learning too. These are guides and studies we read and try to follow, so that what students practise here fits what they meet in school, and later at work.

Swedish schools and workplaces

Skolverket, Technology syllabus: programming to control objects from year 1; own constructions with sensors and electronics in years 7 to 9.
Education Act (chapter 6) and Discrimination Act: active work against degrading treatment, discrimination and harassment.
Arbetsmiljöverket, AFS 2023:2: at work, victimisation is not accepted, and there are routines to handle it.

European guides

DigComp 2.2 (2022): digital competence in five areas: information and data, communication and collaboration, digital content creation, safety, and problem solving.
EU–OECD AI Literacy Framework: engage with AI, create with AI, manage AI and shape AI.
The EU AI Act, article 4 (from February 2025): organisations that use AI must support the AI literacy of their people.

Guides from around the world

UNESCO, AI competency framework for students (2024): a human-centred mindset, the ethics of AI, AI techniques and applications, and AI system design.
TIOBE index (October 2026): C and C++ are among the three most used programming languages.

Studies

Learning by making things that matter to you (Papert, 1980); guided discovery works better than discovery alone (Alfieri and others, 2011).
Teams learn when it is safe to ask and admit mistakes (Edmondson, 1999); humble students try harder after a failure (Porter and others, 2020); feeling left out makes students give up (Walton and Cohen, 2007).
More on why we build real things and in our agreement.

Last looked at: November 2026.