# Autonomiczna jazda

EN: **Autonomous driving** · id `autodriving` · aliases: self-driving cars, autonomous vehicles, autonomous driving systems, samochody autonomiczne, pojazdy autonomiczne, auta bez kierowcy, robotaxi

Path: [AI Atlas](https://aiatlasgraph.com/llms.txt) › Autonomiczna jazda

## Description

- PL: Dziedzina łącząca percepcję, mapy, prognozowanie ruchu, planowanie i symulację, aby pojazd samodzielnie prowadził się w ruchu drogowym.
- EN: The field combining perception, maps, motion forecasting, planning and simulation so that a vehicle drives itself in road traffic.

## In one sentence

- PL: Samochody, które same widzą drogę, przewidują ruch innych i decydują, jak jechać, bez człowieka za kierownicą.
- EN: Cars that see the road, anticipate what others will do and decide how to drive, with nobody at the wheel.

## Problem it solves

- PL: Ludzie za kierownicą męczą się, rozpraszają i popełniają błędy, a wielu w ogóle nie może prowadzić.
- EN: People behind the wheel get tired, distracted and make mistakes, and many cannot drive at all.

## What it does

- PL: Działa jak kierowca z oczami dookoła głowy: łączy sygnały z wielu czujników, zgaduje zamiary innych i co chwilę poprawia plan.
- EN: It works like a driver with eyes all around their head: it merges many sensors, guesses others' intentions and keeps revising its plan.

## Everyday example

- PL: Taksówka bez kierowcy podjeżdża pod dom, omija rowerzystę na pasie i zatrzymuje się, gdy pieszy wchodzi na przejście.
- EN: A driverless taxi pulls up at your door, steers around a cyclist in the lane and stops when someone steps onto the crossing.

## This concerns you if you…

- PL: jeździsz autem, które samo trzyma pas lub odstęp
- PL: zastanawiasz się, kiedy auta pojadą zupełnie same
- PL: pracujesz przy robotach, dronach lub transporcie
- EN: drive a car that keeps its lane or distance by itself
- EN: wonder when cars will drive entirely on their own
- EN: work with robots, drones or transport

## Keywords

SAE levels · perception · sensor fusion · BEV · HD maps · motion forecasting · motion planning · end-to-end driving · CARLA · nuScenes · Waymo Open Dataset · nuPlan · closed-loop evaluation · SOTIF · robotaxi

## How it works

- PL: Klasyczny stos jest modułowy: percepcja z kamer, LiDAR-u i radaru buduje obraz sceny, lokalizacja osadza go na mapie, predykcja szacuje ruch innych, a planer wybiera trajektorię realizowaną przez sterowanie. Coraz częściej moduły łączy się w jedną sieć uczoną end-to-end i ocenia w symulacji w pętli zamkniętej.
- EN: The classic stack is modular: perception from cameras, LiDAR and radar builds a scene picture, localisation places it on a map, prediction estimates others' motion and a planner chooses a trajectory that control executes. Increasingly the modules become one network trained end to end and evaluated in closed-loop simulation.

## Limits

- PL: O bezpieczeństwie decydują rzadkie, nietypowe sytuacje (długi ogon), które trudno zebrać i przetestować. Metryki otwartej pętli słabo przewidują jazdę w ruchu; walidacja wymaga ogromnych przebiegów, symulacji i norm bezpieczeństwa.
- EN: Safety hinges on rare, unusual situations (the long tail) that are hard to collect and test. Open-loop metrics poorly predict on-road driving; validation needs huge mileage, simulation and safety standards.

## Key terms

- PL: **stos modułowy**: Osobne moduły percepcji, predykcji i planowania połączone kolejno.
- PL: **end-to-end**: Jedna sieć zamienia dane z czujników bezpośrednio w plan jazdy.
- PL: **pętla zamknięta**: Test, w którym decyzje modelu zmieniają dalszy przebieg symulacji.
- PL: **poziomy automatyzacji SAE**: Skala 0–5 opisująca, jaką część prowadzenia przejmuje system.
- EN: **modular stack**: Separate perception, prediction and planning modules chained one after another.
- EN: **end-to-end**: One network turns sensor data directly into a driving plan.
- EN: **closed loop**: A test where the model's decisions change how the simulation unfolds.
- EN: **SAE automation levels**: A 0–5 scale describing how much of the driving the system takes over.

## Subtopics (6 direct, 21 in total)

- [Jazda end-to-end (End-to-end autonomous driving)](https://aiatlasgraph.com/t/e2edriving.md): An approach where perception, prediction and planning form one network trained jointly, from sensor data to trajectory.
  - [Modele wizja–język–akcja za kierownicą (Vision-language-action models for driving)](https://aiatlasgraph.com/t/drivingvlm.md): Multimodal models adapted to driving that use images and instructions to predict trajectories, objects or road elements as text or action tokens.
  - [Wspólne uczenie percepcji, predykcji i planowania (Joint perception, prediction and planning)](https://aiatlasgraph.com/t/planningoriented.md): End-to-end architectures with explicit stages (tracking, mapping, prediction, occupancy, planning) linked by queries and trained jointly for planning.
- [Mapy i lokalizacja pojazdu (Maps and localisation for autonomous driving)](https://aiatlasgraph.com/t/admapping.md): Lane-accurate representation of the road network and estimation of the vehicle's pose relative to it from onboard sensors.
  - [Lokalizacja pojazdu na mapie (Map-based vehicle localisation)](https://aiatlasgraph.com/t/maplocalization.md): Estimating the vehicle's position and orientation by fusing GNSS, inertial sensors, odometry and matching observations to a reference map.
  - [Mapy budowane w locie z czujników (Online vectorised HD map construction)](https://aiatlasgraph.com/t/onlinemapping.md): Predicting vector map elements (lane lines, road boundaries, crossings) in a top-down view directly from the vehicle's sensors.
- [Percepcja otoczenia pojazdu (Perception for autonomous driving)](https://aiatlasgraph.com/t/adperception.md): Detecting, tracking and describing objects and free space around the vehicle in 3D from cameras, LiDAR and radar.
  - [Fuzja kamer, LiDAR-u i radaru (Camera–LiDAR–radar sensor fusion)](https://aiatlasgraph.com/t/sensorfusion.md): Combining camera, LiDAR and radar data into one scene representation, often in a shared bird's-eye view (BEV).
  - [Siatki zajętości 3D wokół pojazdu (3D occupancy prediction)](https://aiatlasgraph.com/t/occupancy.md): Predicting for every voxel around the vehicle whether it is occupied and which semantic class it has, instead of object boxes alone.
  - [Widok z lotu ptaka z wielu kamer (BEV) (Bird's-eye-view perception from multi-camera images)](https://aiatlasgraph.com/t/bevperception.md): Transforming features from multiple camera images into a shared top-down grid on which detection, segmentation and planning operate.
- [Planowanie ruchu i sterowanie pojazdem (Motion planning and vehicle control)](https://aiatlasgraph.com/t/adplanning.md): Choosing the vehicle's manoeuvre and trajectory with respect to obstacles, rules and comfort, and executing it with a controller.
  - [Klasyczne planowanie trajektorii (Classical trajectory planning)](https://aiatlasgraph.com/t/rulebasedplanning.md): Planning a driving path with an explicit motion model, search, sampling or optimisation and a hand-defined cost function.
  - [Planery uczone na nagraniach jazdy (Learning-based motion planning)](https://aiatlasgraph.com/t/learnedplanning.md): Trajectory planners trained by imitating human driving, often with extra safety losses and perturbed data.
- [Przewidywanie ruchu innych uczestników ruchu (Motion forecasting for road agents)](https://aiatlasgraph.com/t/motionforecasting.md): Predicting road users' future trajectories a few seconds ahead from motion history, the map and interactions; a model may return several alternatives.
  - [Kodowanie sceny drogowej jako grafu wektorów (Vectorised scene encoding)](https://aiatlasgraph.com/t/vectorizedscene.md): Representing lanes, markings and agent tracks as vector polylines processed by a hierarchical graph network or transformer.
  - [Przewidywanie kilku wariantów ruchu naraz (Multi-modal trajectory prediction)](https://aiatlasgraph.com/t/multihypothesis.md): Generating a set of distinct trajectories with probabilities for each agent, covering different intentions rather than their average.
- [Symulacja i walidacja bezpieczeństwa jazdy (Driving simulation and safety validation)](https://aiatlasgraph.com/t/adsimulation.md): Simulation environments, metrics and processes (scenario testing, standards, on-road statistics) for assessing driving safety and quality.
  - [Generatywne modele świata do jazdy (Generative world models for driving)](https://aiatlasgraph.com/t/drivingworldmodels.md): Generative models predicting future video frames or sensor states, conditioned on the vehicle's actions and text descriptions.
  - [Odtwarzanie scen z nagrań do symulacji czujników (Neural sensor simulation from driving logs)](https://aiatlasgraph.com/t/neuralsensorsim.md): Reconstructing scenes from driving logs with neural representations and rendering camera and LiDAR data from new viewpoints for closed-loop simulation.
  - [Symulacja i benchmarki planowania jazdy (Driving simulation and planning benchmarks)](https://aiatlasgraph.com/t/closedloopsim.md): Environments and benchmarks that evaluate planners either in reactive simulation or on logs with a simplified model of decision outcomes.
  - [Testy scenariuszowe i normy bezpieczeństwa (SOTIF) (Scenario-based testing and safety standards (SOTIF))](https://aiatlasgraph.com/t/scenariotesting.md): Methods and standards for building evidence about driving-system safety: defining operating conditions, scenario tests, analysing functional insufficiencies and using on-road data.

## Sources

- [Self-Driving Cars Specialization (University of Toronto, Coursera)](<https://www.coursera.org/specializations/self-driving-cars>) `coursera-sdc-toronto` · by Steven Waslander, Jonathan Kelly / University of Toronto
- [End-to-end Autonomous Driving: Challenges and Frontiers](<https://arxiv.org/abs/2306.16927>) `e2e-ad-survey-2023` · by Li Chen, Penghao Wu, Kashyap Chitta, Bernhard Jaeger, Andreas Geiger, Hongyang Li
- [Comparison of Waymo Rider-Only Crash Data to Human Benchmarks at 7.1 Million Miles](<https://arxiv.org/abs/2312.12675>) `waymo-crash-7m-2023` · by Kristofer D. Kusano, John M. Scanlon, Yin-Hsiu Chen, Timothy L. McMurry, Ruoshu Chen, Tilia Gode, Trent Victor (Waymo) · sponsored

---

AI Atlas revision 124 (2026-10-31) · HTML: https://aiatlasgraph.com/t/autodriving · index: https://aiatlasgraph.com/llms.txt · everything: https://aiatlasgraph.com/llms-full.txt · updated 2026-10-30