Autonomous Ground Vehicles

An Autonomous Ground Vehicle (AGV), also known as an Unmanned Ground Vehicle (UGV), is defined as a vehicle that operates in contact with the ground without human presence on board. These vehicles are classified into two main categories: those that can be controlled remotely by an operator (remotely controlled) and those that can perceive their environment and make decisions without human intervention to accomplish their tasks (autonomous). With technological advancement, particularly over the last thirty years, they have evolved from simple radio-controlled machines to artificial intelligence-enabled autonomous systems. The functionality of an autonomous vehicle requires the integration of numerous technologies such as self-activating and self-regulating sensors, software, and control systems. These systems are increasingly important in both military and civilian domains.

History

The earliest military applications of unmanned ground vehicles date back to the 19th century. However, modern developments began in the 20th century. In 1928, Elmer E. Wikersham obtained a patent for a UGV designed to deliver explosives to a target. Although this design remained at the prototype stage, it reflected efforts to find alternative solutions for trench warfare.

In the 1930s, the Soviet Union developed the Teletank, a radio-controlled vehicle equipped with a machine gun and operated from another tank. These vehicles were used during the Winter War against Finland (1939–1940) and on the Eastern Front of World War II. During World War II, the British developed a radio-controlled version of the Matilda II infantry tank, known as the "Black Prince." This tank was designed to trigger hidden anti-tank weapons or carry out demolition missions. In 1942, the German army employed a tracked UGV called GOLIATH for demolition operations. This small vehicle, guided by a control cable, could carry up to 100 kg of explosives. However, due to high production costs, low speed (approximately 11.5 km/h), and weak armor, it failed to achieve the expected success in tasks such as tank destruction.

Research on truly autonomous vehicles began in the late 1960s with the SHAKEY project, funded by the Defense Advanced Research Projects Agency (DARPA). SHAKEY, developed at the Stanford Research Institute, was a mobile robot equipped with a TV camera, ultrasonic distance sensors, and tactile sensors, connected via radio to a central computer to perform navigation and reconnaissance tasks. In 1983, DARPA launched the Autonomous Land Vehicle (ALV) program, and vehicles developed under this program reached a speed of 80 km/h by 1990. In the mid-1980s, the United States Marine Corps developed the Teleoperated Vehicle (TOV). During the Gulf War (1990–1991), some M-60 tanks were converted into telerobots for mine-clearing missions. In the 2000s, Honda’s humanoid robot ASIMO made significant contributions to field knowledge.

Technology and Operating Principle

Autonomous ground vehicles rely on a complex technological infrastructure to perceive their environment, process data, and move. This process can be examined in three main stages: perception, planning, and control.

Perception: Sensors that enable the vehicle to "see" its surroundings form the foundation of this stage. These sensors include:

• Lidar (Light Detection and Ranging): Emits laser beams and analyzes reflections to create a precise three-dimensional map of the environment.
• Radar (Radio Detection and Ranging): Uses radio waves to detect the distance, speed, and direction of objects. Particularly effective in adverse weather conditions.
• Cameras: Provide visual data to recognize lanes, traffic signs, pedestrians, and other vehicles.
• Ultrasonic Sensors: Use sound waves to detect nearby obstacles and are commonly used during parking maneuvers.
• Inertial Measurement Unit (IMU): Measures the vehicle’s orientation, speed, and forces acting upon it to assist in position tracking.
• GPS/GNSS: Determines the vehicle’s geographic location through global positioning systems.

Planning and Decision-Making: Data collected by sensors is combined and analyzed in the vehicle’s central processing unit. In this stage, artificial intelligence algorithms are activated. The vehicle uses this data to detect static and moving obstacles, predict road shape and lane markings, and calculate its own position, speed, and status. Based on this information, it plans a safe route and makes decisions such as overtaking, lane changing, or stopping. In military applications, artificial intelligence can also perform tasks such as target detection, classification, and prioritization.

Control and Movement: Decisions made during the planning stage are transmitted to the vehicle’s mechatronic units. These units control steering, acceleration, and braking systems to enable physical movement. This process allows the driver or operator to monitor and intervene in the system through a human-machine interface (HMI) when necessary.

Levels of Autonomy

Autonomous driving technology is classified according to a six-level standard established by the Society of Automotive Engineers (SAE). These levels define the degree of automation and the driver’s responsibilities.

• Level 0 (No Driving Automation): The driver is fully responsible for all driving tasks. No automated systems are present in the vehicle.
• Level 1 (Driver Assistance): The system can assist with only one driving task, such as steering or acceleration/deceleration. Adaptive cruise control is an example of this level. The driver performs all other tasks.
• Level 2 (Partial Driving Automation): The system can simultaneously handle both steering and acceleration/deceleration. However, the driver must continuously monitor the driving environment and be ready to take control at any moment.
• Level 3 (Conditional Driving Automation): The vehicle can assume all driving tasks under specific conditions (e.g., traffic congestion on a highway). The driver is not required to continuously monitor the road, but must be ready to take control when requested by the system.
• Level 4 (High Driving Automation): The vehicle can perform all driving tasks without human intervention within a specific geographic area or operational context (e.g., campus shuttle service). The system may not function outside these conditions.
• Level 5 (Full Driving Automation): The vehicle can perform all driving tasks entirely on its own under any road and environmental conditions a human driver could handle. No human driver is required at this level.

Applications

Autonomous ground vehicles are used across a wide range of military and civilian applications.

Military Applications

In the military domain, UGVs are employed to ensure personnel safety during hazardous missions and to provide a force multiplier effect. Their primary roles include:

• Reconnaissance, Surveillance, and Intelligence (RSI): Gathering information beyond enemy lines or in dangerous areas.
• Fire Support: Neutralizing enemy targets using integrated weapon systems. In such systems, the decision to fire is typically made by an operator following the "human-in-the-loop" principle.
• Logistical Support: Transporting ammunition, supplies, and equipment to units on the battlefield.
• Mine and Improvised Explosive Device (IED) Detection/Neutralization: Conducting safe clearance operations in high-risk zones.
• Other Missions: Can be used for a variety of tasks such as tactical deception, fortification reconnaissance, communication relaying, and casualty evacuation.

Various UGVs have been developed by defense industry companies in Türkiye.

Some of these include:

• Gölge Süvari: A modular, multi-purpose UGV family developed by FNSS on the M113 platform. It can autonomously perform missions such as fire support, reconnaissance, and logistics.
• BARKAN: A multi-purpose vehicle capable of integrating various payloads such as weapon systems or robotic arms, featuring autonomous patrol and leader-following capabilities.
• ASLAN: A UGV capable of reconnaissance, surveillance, and target detection, operable either autonomously or via remote command.
• HANÇER: A medium-class UGV designed for challenging terrain conditions, equipped with AI-supported target recognition and prioritization capabilities.
• FEDAİ: A light UGV developed specifically for urban operations, advancing ahead of personnel to conduct reconnaissance and neutralize threats.

Civilian Applications

In civilian contexts, autonomous technologies hold the potential to revolutionize transportation, logistics, and industrial automation. Key advantages of autonomous vehicles in daily life include reducing traffic accidents, enhancing mobility for the elderly and disabled, and optimizing traffic flow. Technologies such as automated valet parking systems, highway pilots, and traffic congestion assistants represent the initial steps toward fully autonomous systems. Additionally, autonomous robots are being developed for specialized industrial tasks such as gas leak detection or hazardous material transport.

Legal and Ethical Issues

The widespread adoption of autonomous vehicles brings significant legal and ethical challenges. The most fundamental debate concerns liability in the event of an accident. Responsibility may be shared among the vehicle’s manufacturer, the software developer, the vehicle owner, the current user (driver), or regulatory authorities. It is also debated whether the vehicle itself could be recognized as a legal person and held criminally liable; however, under existing legal systems, machines are generally not considered capable of intent or negligence, making this approach largely unfeasible. Countries such as Germany, the United States, Japan, and Türkiye are working on regulatory frameworks to address the legal status of autonomous vehicles.