LO
Driving instructor — tasks, one by one
The unit of analysis is the task, not the job title. Each one below carries its direction, whether the judgement rests on evidence or on platform inference, the reasoning, and what it does not establish.
Every task on this page#
Basic control on the training ground
Automating✓ Evidence-backedTeaching starting, steering, reversing and parking manoeuvres on a closed training ground before a pupil goes on the road.
This is where regulators are testing teaching without an instructor in the seat. Shanghai's transport commission defines in-car AI teaching equipment for the basic and training-ground part of the syllabus that partly or fully takes the place of the instructor's in-car guidance, lets pupils train alone in such cars on a separate closed ground, and requires on-site instructors equal to only 20% of those cars, at least 2. Guangdong's transport department runs a pilot of intelligent robot coach cars for the same part of the syllabus. Both keep qualified instructors watching from a monitoring post or on the ground.
Two regional pilots in one country, open to a few selected schools; neither reports results, and neither covers driving on public roads.
Lessons in real traffic
Still human-led✓ Evidence-backedSitting beside a learner on public roads, judging when to step in, and using the dual controls when needed.
Every rule found keeps a person in the car on the road. China's national syllabus allows driving simulators but caps simulator teaching at 6 hours; Shanghai confines its AI teaching equipment to a closed ground and forbids mixing those cars with others. A systematic review of novice-driver education found simulators improve skills such as hazard anticipation, but that the gains do not always translate into lower real-world crash risk without continued training or real-life practice.
Rules and a review of studies; no record measures lessons on public roads being taught by a machine, and rules can change.
Theory, hazard awareness and car technology
Being augmented✓ Evidence-backedTeaching the rules of the road, how to spot hazards, and how driver-assistance features work and where they stop working.
Technology takes part of this teaching and adds new material to the rest. China's syllabus encourages combining classroom and online teaching and adds driver-assistance features to what must be taught. The EU's 2025 driving licence directive requires theory tests to ask about the advantages, limitations and risks of driver assistance and automated driving systems, and lets hazard skills be tested on simulators.
Rules about what must be taught and tested; they do not say how much of it instructors, apps or classrooms deliver.
Assessing progress and preparing for the test
Still human-led✓ Evidence-backedJudging when a pupil is ready, rehearsing test routes and manoeuvres, and explaining what the examiner will look for.
The test still assesses a person driving. The UK driving test agency allows cruise control, parking sensors and cameras, lane assist and similar features during the test, but says self-parking and Tesla Autopilot cannot be used, and that the rules also apply to the cars instructors provide for tests.
One country's test rules; they say what a pupil must do unaided, not how instructors assess readiness.
Bookings and running the business
Being augmented≈ Platform inferenceScheduling lessons, booking tests, taking payments and finding new pupils.
No record found here. Booking and payment are the kind of routine work that scheduling software already handles for many small businesses.
The absence of a record means this judgement rests on how similar work has changed elsewhere, not on evidence about driving instructors.