China Tests AI Patrol Boat, Can Approach Victims in Water in 3 Minutes

The Haixun 14102 ship operates in Shenzhen waters. This maritime patrol ship has been upgraded digitally and AI-based to support surveillance and rescue operations. Photo: China Daily.

China has begun testing an artificial intelligence (AI) based maritime patrol ship in Shenzhen, Guangdong Province. The ship can follow routes, avoid obstacles, recognize objects, and help simulate rescues without having to be constantly steered by officers.

China Daily, quoted Tuesday, August 4, reported that the 12-meter-long ship named Haixun 14102 began undergoing trial operations in early July. This ship is said to be the first maritime law enforcement ship in China that has been digitally and intelligently upgraded.

Haixun 14102 is equipped with an intelligent navigation module, a number of sensors, and a platform that connects the ship to a control center on land.

In one test, the ship traveled 14.2 nautical miles. The deviation from the planned route remained below 1 meter. In a rescue simulation, the ship's system detects, recognizes, and approaches people in the water within three minutes.

The ship can also be operated remotely under the supervision of ground personnel.

The technology is not designed to replace officers. The goal is to help officers read the situation faster, especially during patrols and operations to search for victims in busy waters.

"The search and rescue operation could last 36 or even 48 hours," said Hao Xianlin, head of the airport maritime patrol unit under the Bao'an Maritime Safety Administration.

"An integrated sensing system can greatly improve our ability to detect and identify the people we are looking for," he said.

Haixun 14102 has a maximum speed of 36 knots, a speed unit commonly used at sea. The ship was chosen for the trial because it is widely used for routine patrols, emergency response, and rescue operations.

Shenzhen is in the Guangdong-Hong Kong-Macao Greater Bay Area. In this region, commercial vessels, passenger ferries, and small boats move on crowded waterways.

For maritime officers, the challenge is not just the number of ships. Hours of visual observation can trigger fatigue. Dark conditions, fog, rough seas, and heavy traffic also make patrols more difficult.

Video and sensor data from Haixun 14102 are sent to the command center on land. Officers can monitor the surrounding conditions directly and provide operational support if needed.

This project is also considered more economical than building a smart ship from scratch. Engineers added navigation systems, sensors, communications, and ship-to-land connections without changing the ship's main structure or lowering its original performance.

Shenzhen maritime authorities said the increase in old ships could cut costs and speed up the adoption of technology.

This trial is part of China's push to build smart shipping. The country has one of the largest water transport networks in the world.

In 2025, Chinese ports handled 18.34 billion metric tons of cargo and 354 million TEU containers. A TEU is a standard unit for calculating 20-foot containers. China's navigable inland waterways reach 128,700 kilometers, according to the Chinese Ministry of Transport.

China Daily noted that China has also prepared a Smart Shipping 2030 Action Plan. The plan encourages the integration of AI with ships, ports, waterways, navigation services, and maritime surveillance.

By 2027, China aims to have at least three smart shipping pilot areas, more than five pilot routes, more than 10 replicable application scenarios, and more than 100 smart ships in operation.

Priority applications include emergency rescue, night or fog navigation, operations in narrow and winding lanes, construction work in waters, and short-distance transportation.

However, experts warn that smart shipping cannot be seen as simple as attaching sensors or eliminating crew members.

"No matter how sophisticated the laboratory simulation is, it cannot reproduce every wave, current, or unexpected situation in the sea," said Yin Yong, a professor at the Navigation College, Dalian Maritime University.

According to Yin, the most practical model today is cooperation between ships and land operators. In open waters, ships can rely more on autonomous navigation. However, in narrow lanes or bad weather, land operators can still intervene.

"It allows us to use the advantages of AI while still keeping human judgment available when it's needed most," Yin said.

Similar technology is also beginning to be expanded to navigation infrastructure. In the Port of Jiangyin, Fujian, navigation aids are equipped with weather monitors and Automatic Identification System or AIS devices.

AIS is an automatic identification system that helps monitor the position and information of ships. In Jiangyin, drones are used for routine inspections, while connected buoys give warnings about coral and other hazards.

In Minjiang River, Fujian, tourist boats are monitored throughout the night voyage via a digital maritime platform.

On the Minjiang River and the Lijiang River, electric and new energy vessels are beginning to replace old diesel vessels. The goal is to reduce noise, emissions, and the risk of pollution on water tourism routes.

Even so, widespread implementation still needs testing, inspection, certification, operational standards, cyber security protection, and clear division of responsibilities between ship crews and land operators.

Shenzhen maritime authorities are now finalizing training, inspection and certification procedures, law enforcement protocols, and maintenance standards for Haixun 14102. The experience from this trial will determine whether similar improvements can be applied to more maritime patrol and rescue vessels.