Introduction
Autonomous systems are no longer limited to research laboratories. They are already being used in transportation, manufacturing, agriculture, logistics, healthcare, infrastructure inspection, defence, environmental monitoring, space exploration, and household applications.
Autonomous cars, drones, and robots are among the most visible examples of this technology. Although these systems operate in different environments, they follow a similar process. They collect information through sensors, interpret their surroundings, estimate their position, make decisions, plan actions, control movement, and continuously evaluate the results.
An autonomous car must understand roads, traffic signals, pedestrians, and nearby vehicles. A drone must maintain stable flight, avoid obstacles, and respond to wind and changing environmental conditions. A robot may need to navigate through a warehouse, identify objects, perform physical tasks, or cooperate safely with people.
These examples demonstrate that autonomy is not created by a single artificial-intelligence model. It results from the integration of sensors, perception, machine learning, localization, planning, control, actuators, communication, and safety mechanisms.
Autonomous Cars
An autonomous car is a vehicle capable of performing some or all driving functions with reduced human involvement. Depending on its level of automation, it may assist the driver, control selected driving functions, or perform the complete driving task within defined operating conditions.
The main objective of an autonomous car is to transport passengers or goods safely from one location to another. To achieve this, the vehicle must understand the road environment, follow traffic rules, predict the behaviour of other road users, and control its movement in real time.
Autonomous cars commonly use cameras, radar, LiDAR, ultrasonic sensors, satellite-navigation systems, inertial measurement units, and wheel sensors.
Cameras provide detailed visual information about road markings, traffic lights, signs, pedestrians, vehicles, and surrounding objects. Radar measures the distance and speed of moving objects and can remain useful during poor visibility. LiDAR creates a three-dimensional representation of the surrounding environment. Ultrasonic sensors support short-range detection during parking and low-speed movement.
Satellite navigation helps determine the vehicle’s geographic position, while inertial sensors and wheel measurements estimate movement when satellite signals are inaccurate or temporarily unavailable.
The vehicle combines information from these sensors through sensor fusion. This process creates a more reliable understanding of the environment than any individual sensor can provide.
The perception system identifies road lanes, vehicles, cyclists, pedestrians, signs, traffic signals, barriers, and other objects. Tracking algorithms estimate how these objects move over time. Prediction models evaluate what they may do next.
For example, the vehicle may identify a pedestrian standing near the edge of a road. It must determine whether the pedestrian is stationary, walking away, or preparing to cross.
The decision-making system then selects an appropriate response. It may continue at the same speed, reduce speed, stop, change lanes, or increase the distance from the pedestrian.
The planning system generates a safe path and trajectory. The control system converts the trajectory into steering, acceleration, and braking commands.
This process operates continuously because the road environment changes every moment.
Lane Keeping and Adaptive Speed Control
Lane-keeping systems use cameras to identify road markings and estimate the vehicle’s position within the lane. If the vehicle moves too close to the lane boundary, the system may issue a warning or adjust the steering.
Adaptive speed-control systems use radar or cameras to detect the vehicle ahead. The system maintains a selected speed when the road is clear and reduces speed when another vehicle is closer than the safe following distance.
These technologies provide important automated functions, but they do not necessarily make the vehicle fully autonomous. The human driver may still remain responsible for monitoring the road and controlling other aspects of driving.
Automatic Emergency Braking
Automatic emergency braking is a safety function that monitors the road ahead for collision risks.
If the system detects a vehicle, pedestrian, cyclist, or obstacle and determines that a collision may occur, it warns the driver. If the driver does not respond quickly enough, the system may apply the brakes automatically.
This function demonstrates how perception, prediction, decision-making, and actuation work together.
The sensors identify an object, the system estimates distance and relative speed, a risk model calculates the likelihood of collision, and the controller activates the brakes when necessary.
Autonomous Parking
Autonomous parking systems assist or control the process of moving a vehicle into a parking space.
The system may use cameras and ultrasonic sensors to identify available space and measure the distance from nearby vehicles, walls, and obstacles.
It then calculates a suitable path and controls steering. More advanced systems may also control acceleration, braking, and gear selection.
Parking is generally easier to automate than driving in open traffic because the vehicle moves slowly and operates within a smaller area. However, the system must still detect pedestrians, shopping carts, pillars, and unexpected movement.
Highway Driving
Highways provide a more structured environment than many urban roads. Traffic normally moves in the same direction, lanes are clearly marked, and there are fewer pedestrians and intersections.
Autonomous highway systems may maintain lane position, adjust speed, follow another vehicle, perform lane changes, and respond to traffic conditions.
Before changing lanes, the system must detect nearby vehicles, estimate their speed, calculate available space, predict future positions, and determine whether the manoeuvre is safe.
Although highways are structured, challenges remain. Construction zones, faded road markings, unusual vehicles, debris, heavy rain, and sudden human behaviour can affect performance.
Urban Autonomous Driving
Urban driving is one of the most difficult autonomous applications.
The vehicle must understand traffic signals, intersections, pedestrians, bicycles, parked cars, buses, emergency vehicles, road signs, lane changes, and complex right-of-way rules.
Urban environments also contain partially hidden objects and unpredictable behaviour. A child may run into the road, a cyclist may change direction suddenly, or a parked vehicle may begin moving.
The autonomous system must combine perception, prediction, rules, planning, and safety monitoring while making decisions within a very short time.
For this reason, many autonomous vehicles operate only within restricted urban zones, mapped areas, limited speeds, or selected weather conditions.
Autonomous Shuttles
Autonomous shuttles are often used on university campuses, airports, business parks, hospitals, and controlled public routes.
They usually operate at low speeds and follow predefined or mapped paths. The operating environment is restricted, making perception and planning more manageable.
The shuttle detects pedestrians, vehicles, obstacles, and stopping points. It may communicate with a remote monitoring centre and enter a safe-stop condition if it encounters a situation it cannot manage.
This is an example of high autonomy within a limited Operational Design Domain.
Autonomous Trucks and Logistics Vehicles
Autonomous trucks are being developed for highways, ports, mines, warehouses, and logistics centres.
Long-distance highway transport contains many repetitive driving conditions, making it suitable for advanced automation. Ports and mines provide even more controlled environments where vehicle routes and access can be managed.
Autonomous logistics vehicles may transport containers, raw materials, or goods without a driver. They use digital maps, positioning systems, obstacle detection, route planning, and fleet communication.
Fleet-management software can assign tasks and coordinate several vehicles to prevent congestion.
The system must also monitor fuel or battery level, mechanical condition, tyre pressure, communication quality, and route status.
Challenges in Autonomous Cars
Autonomous cars must perform safely in complex and changing conditions. Poor weather, fog, rain, glare, darkness, damaged roads, unclear signs, and unusual objects can reduce sensor performance.
Human behaviour is another major challenge. Drivers, pedestrians, and cyclists may behave unpredictably or ignore traffic rules.
The system must also manage rare situations that may not have appeared sufficiently in its training data.
Cybersecurity is important because connected vehicles communicate with external systems and may receive software or map updates.
Legal responsibility, insurance, privacy, testing standards, and public trust also affect deployment.
An autonomous car must therefore be evaluated as a complete safety-critical system rather than only as an artificial-intelligence application.
Autonomous Drones
An autonomous drone is an unmanned aerial system capable of performing flight and mission tasks with limited direct human control.
A manually controlled drone depends on a pilot for direction, altitude, and movement. An autonomous drone may follow a planned route, maintain stable flight, avoid obstacles, collect data, return to its starting point, and respond to failures independently.
Drones operate in three-dimensional space, which makes their movement more complex than that of many ground robots. They must manage altitude, direction, speed, rotation, stability, wind, and energy use.
Autonomous drones commonly use cameras, GPS, inertial measurement units, magnetometers, altitude sensors, LiDAR, radar, and ultrasonic sensors.
The inertial measurement unit measures acceleration and rotation. GPS provides geographic position. Altitude sensors estimate height, while cameras or LiDAR help detect obstacles and map the environment.
The flight controller continuously adjusts motor speed to maintain stability. Even when the drone appears stationary, the controller makes constant corrections to compensate for wind and movement.
Route-Based Autonomous Flight
A route-based drone follows a series of predefined geographic points known as waypoints.
The operator defines the mission before launch. The drone then calculates how to travel between the points, maintain altitude, collect data, and return safely.
Waypoint navigation is used in mapping, agriculture, inspection, surveying, and environmental monitoring.
The drone may adjust its speed according to wind, battery level, or image-capture requirements.
If GPS accuracy decreases or an obstacle appears, the drone may pause, change its route, or request assistance.
Agricultural Drones
Agricultural drones are used to monitor crops, analyse plant health, map fields, estimate irrigation needs, and support precision farming.
A drone may follow an automated route across a field while capturing visible, thermal, or multispectral images.
Artificial-intelligence models can analyse the images to identify crop stress, disease, pest damage, water shortage, nutrient deficiency, or weed growth.
The drone may generate a map showing areas that require attention. Farmers can then apply water, fertilizer, or treatment more precisely.
More advanced agricultural drones may also perform spraying or seeding operations.
The drone must control altitude and speed carefully to maintain consistent image quality or spray coverage. Wind, trees, power lines, and changing terrain must also be considered.
Infrastructure Inspection Drones
Drones are increasingly used to inspect bridges, towers, pipelines, wind turbines, solar panels, roofs, and power lines.
Traditional inspection may require workers to climb structures or use expensive equipment. Autonomous drones can reduce risk by capturing images and sensor data from difficult locations.
The drone may follow a predefined inspection path around the structure. Computer vision can identify cracks, corrosion, missing components, heat differences, or other defects.
Autonomous control helps maintain a stable distance from the structure. This is important because the drone must collect detailed images without colliding with the surface.
GPS may be unreliable near large structures, so the drone may rely on cameras, LiDAR, inertial sensing, and visual localization.
Delivery Drones
Delivery drones are designed to transport small packages, medicines, food, or emergency supplies.
The system receives a delivery destination and calculates a flight route. It must consider restricted airspace, buildings, weather, battery level, and safe landing locations.
During flight, the drone monitors its position and detects obstacles. It may communicate with a fleet-control platform that tracks several deliveries.
The final stage of delivery can be difficult. The drone must identify the correct location, ensure that the landing area is clear, and release the package safely.
Delivery drones may be especially useful in remote areas, during disasters, or when urgent medical supplies are required.
Search-and-Rescue Drones
Search-and-rescue drones can support emergency teams during floods, earthquakes, fires, avalanches, and missing-person operations.
They can cover large areas quickly and access locations that may be unsafe for human rescuers.
Cameras, thermal sensors, and AI-based image analysis may help detect people, movement, heat signatures, damaged buildings, or blocked routes.
An autonomous drone may divide an area into search zones and follow a systematic route. If a possible person is detected, it can send coordinates and images to the rescue team.
The system must handle poor visibility, smoke, wind, damaged infrastructure, and weak communication.
Human verification remains important because incorrect detection can waste critical rescue time.
Environmental-Monitoring Drones
Drones can monitor forests, rivers, coastlines, wildlife, air quality, and environmental change.
They may collect images, temperature measurements, gas readings, water information, or vegetation data.
Autonomous flight allows repeated monitoring of the same area over time. This supports comparison and change detection.
AI models may identify illegal logging, forest fires, pollution, erosion, damaged habitats, or changes in animal populations.
These applications demonstrate that autonomy can improve data collection in large, remote, or difficult environments.
Indoor Autonomous Drones
Indoor environments create special navigation challenges because GPS signals are normally unavailable.
Indoor drones may use cameras, LiDAR, depth sensors, and SLAM to estimate position and build maps.
They can inspect warehouses, tunnels, factories, mines, and damaged buildings.
The system must avoid walls, ceilings, cables, equipment, and people while maintaining stable flight.
Indoor drones require accurate localization because even a small position error can cause collision in a narrow space.
Drone Safety and Fallback Behaviour
Autonomous drones must monitor battery level, motor condition, communication quality, wind, position accuracy, and sensor status.
When the battery becomes low, the drone may return to its launch point or land at a safe location.
If communication is lost, it may continue the mission, remain in place, return home, or land, depending on the programmed safety policy.
If GPS becomes unreliable, the drone may switch to visual navigation or reduce movement.
Geofencing prevents the drone from entering restricted areas. Emergency landing procedures reduce risk when normal flight is no longer possible.
The appropriate fallback behaviour depends on the environment. Returning directly to the launch point may be unsafe if a new obstacle has appeared along the route.
Autonomous Robots
The term robot describes a programmable machine capable of sensing, processing information, and performing physical actions. An autonomous robot can complete selected tasks with reduced continuous human control.
Robots may move through an environment, manipulate objects, interact with people, or operate as stationary intelligent machines.
Autonomous robots are used in warehouses, factories, hospitals, farms, homes, laboratories, mines, oceans, and space.
Different robots require different combinations of perception, planning, control, and physical design.
Warehouse Robots
Warehouse robots transport products, shelves, or containers between storage and packing areas.
They use maps, LiDAR, cameras, wheel encoders, and proximity sensors to navigate.
A central system may assign tasks to each robot. The robot then calculates a route, avoids workers and other robots, and delivers the item.
When several robots operate together, fleet-management software coordinates their movement. It may assign priority, prevent traffic congestion, and redirect robots around blocked pathways.
The robot also monitors its battery and returns to a charging station when required.
Warehouse environments are suitable for high autonomy because pathways, maps, speed limits, and access conditions can be controlled.
However, robots must still respond safely to people, dropped objects, moving equipment, and changing layouts.
Autonomous Mobile Robots
Autonomous mobile robots can navigate without following a permanently installed path.
Traditional automated guided vehicles often follow magnetic strips, wires, markers, or predefined tracks. Autonomous mobile robots use sensors, maps, and navigation algorithms to select their own routes.
They can identify obstacles and calculate alternative paths when the original route becomes blocked.
These robots are used in factories, warehouses, hospitals, offices, and commercial buildings.
A hospital robot may transport medicines, laboratory samples, food, or equipment. It must navigate corridors, avoid patients and staff, use lifts, and reach the correct department.
The robot may communicate its status through a screen, sound, or remote dashboard.
Industrial Robotic Arms
Industrial robotic arms perform tasks such as welding, assembly, painting, packaging, cutting, and material handling.
Traditional robotic arms repeat fixed movements in controlled environments. Intelligent robotic arms use cameras, force sensors, and AI to adapt their actions.
A vision-guided robot can identify parts placed in different positions. It estimates the location and orientation of each item, calculates a grasping motion, and moves the gripper accordingly.
Force sensors help determine whether the robot has held the object securely.
If the object moves or the grasp fails, the system can adjust and try again.
Industrial robots must follow strict safety limits because they may move rapidly and handle heavy equipment.
Collaborative Robots
Collaborative robots, commonly called cobots, are designed to work near or with human workers.
Instead of remaining completely separated behind safety barriers, cobots share a workspace with people.
They may assist with assembly, lifting, inspection, packaging, or tool handling.
Sensors monitor force, speed, distance, and human movement. If a person comes too close, the robot may slow down or stop.
Shared autonomy is common in collaborative robotics. The human provides judgment and flexibility, while the robot provides strength, accuracy, and repeatability.
The design must consider human comfort, trust, communication, and predictable movement.
A robot that changes direction suddenly may frighten workers even when it does not create a physical danger.
Service Robots
Service robots perform useful tasks outside traditional industrial production.
Examples include cleaning robots, delivery robots, hotel robots, restaurant robots, information robots, and security robots.
A robotic vacuum is a familiar household example. It uses sensors to detect walls, furniture, stairs, and obstacles. More advanced models create maps and calculate cleaning routes.
Delivery robots may move along pavements or inside buildings. They must avoid pedestrians, cross suitable pathways, identify delivery locations, and maintain communication with operators.
Service robots often operate in environments designed for humans. This creates challenges because doors, stairs, crowds, pets, uneven surfaces, and social behaviour can be difficult for machines to understand.
Healthcare Robots
Healthcare robots support surgery, rehabilitation, patient monitoring, medicine delivery, and hospital logistics.
A surgical robot may improve precision and filter unwanted hand movement, but important decisions remain under the control of trained medical professionals.
Rehabilitation robots assist patients in performing repeated movements. The system may adjust support according to the patient’s strength and progress.
Hospital delivery robots transport medicines, samples, food, and equipment. They can reduce routine workload and limit unnecessary movement through sensitive areas.
Socially assistive robots may communicate with patients, provide reminders, or support elderly care.
Healthcare autonomy requires strong safety, privacy, reliability, and human oversight because errors can directly affect patient well-being.
Agricultural Robots
Agricultural robots perform tasks such as planting, weeding, harvesting, crop monitoring, and precision spraying.
A weeding robot may use cameras and AI to distinguish crops from unwanted plants. It then removes or treats the weeds without affecting the crop.
A harvesting robot must identify ripe fruit, estimate its position, plan a movement, and apply enough force to collect it without causing damage.
Agricultural environments are difficult because lighting, weather, soil, vegetation, and object appearance change continuously.
Fields may contain uneven terrain, mud, dust, animals, people, and machinery. Robots must therefore combine robust perception with adaptable movement and safety controls.
Mining and Construction Robots
Autonomous vehicles and robots are used in mines and construction areas to transport materials, inspect tunnels, drill, excavate, and monitor hazardous locations.
These environments may contain dust, unstable surfaces, poor lighting, falling material, and limited communication.
Removing human workers from dangerous areas can improve safety.
Autonomous haulage vehicles may follow mapped routes between excavation and unloading locations. Fleet systems coordinate movement and prevent collisions.
Inspection robots can enter narrow tunnels or unstable structures to collect images and environmental data.
The machines must be rugged, reliable, and capable of safe operation when sensors or communication become limited.
Search-and-Rescue Robots
Ground robots can enter collapsed buildings, tunnels, fire zones, contaminated areas, or disaster sites.
They may use cameras, thermal sensors, gas sensors, microphones, and mapping systems to locate people and assess conditions.
The robot may move across rubble, climb stairs, or pass through narrow spaces.
Complete autonomy may be difficult because disaster environments are highly unstructured. Many search-and-rescue robots therefore use shared autonomy.
The robot manages low-level movement and obstacle avoidance, while a human operator provides mission-level decisions.
Underwater Robots
Autonomous underwater vehicles are used for ocean research, pipeline inspection, seabed mapping, environmental monitoring, and offshore operations.
Underwater communication is difficult, and satellite navigation does not work beneath the surface.
The robot may use sonar, inertial navigation, depth sensors, cameras, and acoustic positioning.
It must manage water currents, pressure, limited visibility, energy use, and communication delays.
Because continuous remote control may be impossible, the vehicle must follow mission plans and respond to obstacles or faults independently.
It may return to the surface when the mission is complete or when battery power becomes low.
Space Robots
Planetary rovers and robotic spacecraft require a high degree of autonomy because communication with Earth may be delayed.
A rover must identify rocks, slopes, holes, and unsafe terrain. It estimates its position, selects a path, manages energy, and performs scientific tasks.
Human operators define mission goals, but the robot may need to make immediate navigation decisions without waiting for instructions.
Space systems must be extremely reliable because physical repair may be impossible.
They must also manage radiation, extreme temperatures, rough terrain, limited power, and communication restrictions.
Laboratory Robots
Autonomous laboratory systems are used in chemistry, biology, materials research, and pharmaceutical development.
A laboratory robot may prepare samples, operate instruments, record results, and adjust the next experiment according to previous findings.
Artificial intelligence can analyse experimental data and recommend new conditions.
This creates a closed research cycle in which the system plans an experiment, performs it, measures the result, analyses the data, and selects the next experiment.
Human researchers remain responsible for defining objectives, interpreting significance, ensuring safety, and reviewing outcomes.
Similarities Across Cars, Drones, and Robots
Autonomous cars, drones, and robots may appear very different, but they share the same basic operating principles.
Each system requires sensors to observe the environment and its internal condition. The raw information must be processed, synchronized, and interpreted.
Perception identifies objects, free space, hazards, and environmental features. Localization determines the system’s position and orientation.
Decision-making selects an appropriate behaviour, while planning determines how the behaviour should be performed.
The control system converts the plan into physical commands. Actuators move the vehicle, drone, or robot.
Feedback confirms whether the action was completed correctly. Safety systems monitor faults and may stop or modify the operation.
The systems also share common limitations. Sensor noise, unfamiliar situations, uncertain predictions, hardware faults, communication problems, cybersecurity threats, and human misunderstanding can affect performance.
Differences Between Cars, Drones, and Robots
Autonomous cars mainly operate on roads and must follow traffic laws, lane structures, and social driving behaviour. Their movement is generally limited to a two-dimensional surface, but road traffic is highly dynamic and safety-critical.
Drones move through three-dimensional space. They must maintain flight stability while managing altitude, wind, battery use, and airspace restrictions. Even a temporary motor or control failure can cause the drone to fall.
Robots vary widely in form and purpose. A mobile robot may navigate a building, while a robotic arm remains fixed and manipulates objects. A surgical robot requires different safety controls from a mining robot.
The environment strongly influences system design. A warehouse is structured and controlled, while a public road or disaster site is open and unpredictable.
The acceptable level of autonomy also differs. A cleaning robot may be allowed to make most decisions independently, while a medical robot may require direct human approval for critical actions.
Human Roles in Real-World Autonomy
Autonomous operation does not remove all human involvement.
Humans define the mission, select operating conditions, maintain the equipment, review performance, update software, manage unusual situations, and take responsibility for deployment.
A vehicle may drive independently but still depend on remote assistance. A drone may follow a route autonomously while a pilot supervises the mission. A warehouse robot may move independently after receiving a task from a central system.
In human-in-the-loop systems, a person approves important decisions before action.
In human-on-the-loop systems, the machine performs the task while a person monitors and can intervene.
In human-out-of-the-loop operation, the system makes immediate decisions independently, usually because communication is delayed or unavailable.
The correct level of human involvement depends on the risk, task, environment, and consequences of failure.
Safety in Real-World Autonomous Systems
Safety must be included at every stage of autonomous-system design.
Cars require collision avoidance, emergency braking, safe following distance, reliable steering, and controlled fallback behaviour.
Drones require geofencing, low-battery response, obstacle avoidance, emergency landing, and communication-loss procedures.
Robots require speed limits, force monitoring, restricted zones, emergency stops, and safe interaction with people.
Redundant sensors and processors may be used when a single failure could cause serious harm.
The system must detect when its perception, localization, communication, or actuator performance becomes unreliable.
When safe operation is no longer possible, it should enter a minimum-risk condition. This may involve stopping, landing, returning to a base station, reducing speed, or requesting human assistance.
Ethical and Social Considerations
Real-world autonomy creates questions beyond technical performance.
Autonomous vehicles may affect employment, transport access, insurance, and legal responsibility.
Drones may create privacy concerns when capturing images over homes, farms, public areas, or sensitive locations.
Workplace robots may change job roles and require new skills. Healthcare robots may collect personal and medical data.
Autonomous surveillance systems may affect civil liberties if they are used without appropriate limits.
Developers and organizations must consider fairness, privacy, transparency, accountability, human control, accessibility, and public benefit.
A system should not be deployed only because it is technically possible. Its risks, users, environment, and wider social effects should also be evaluated.
Limitations of Real-World Autonomous Systems
Most autonomous cars, drones, and robots are designed for specific tasks and operating environments.
A warehouse robot may perform reliably indoors but fail on a public pavement. A drone designed for open agricultural fields may not operate safely inside a building. A vehicle designed for clear highway conditions may not manage a flooded urban road.
This limitation is described through the Operational Design Domain.
The system should clearly communicate where, when, and under what conditions its autonomous functions are available.
A machine should not be treated as fully capable simply because it performs well during normal conditions.
Rare events, unexpected objects, sensor failures, poor weather, unfamiliar locations, and unusual human behaviour must also be considered.
Real-world autonomy therefore requires continuous testing, monitoring, maintenance, and improvement.
Conclusion
Autonomous cars, drones, and robots demonstrate how artificial intelligence, sensors, planning, control, and physical action can be combined to perform real-world tasks.
Autonomous cars use cameras, radar, LiDAR, navigation systems, prediction models, planning algorithms, and controllers to understand roads and manage driving functions. Their applications include assisted driving, automated parking, autonomous shuttles, highway operation, and logistics transport.
Autonomous drones use flight controllers, navigation sensors, cameras, obstacle detection, and mission-planning software. They support agriculture, delivery, infrastructure inspection, environmental monitoring, mapping, and emergency response.
Autonomous robots operate in warehouses, factories, hospitals, farms, mines, homes, laboratories, oceans, and space. They transport materials, inspect environments, manipulate objects, assist people, and perform tasks in hazardous locations.
Although these systems have different physical forms, they follow a common cycle of sensing, perceiving, estimating, deciding, planning, controlling, acting, and evaluating.
Their effectiveness depends on accurate environmental understanding, reliable decision-making, stable control, safe actuation, continuous feedback, and clear operating limits.
Most present-day autonomous systems are not completely independent in every environment. They operate within defined conditions and continue to depend on humans for supervision, maintenance, mission definition, safety management, and accountability.
These real-world examples show both the potential and the limitations of autonomy. They also provide the practical foundation for studying sensors, computer vision, machine learning, navigation, control, safety, and responsible autonomous-system development in greater depth.
Comments are closed.