Nuclear Intelli-Inspection Robot Dog: Safety Fortified

I. Project Background

    The reactor building of Unit 3 at a nuclear power plant contains critical equipment such as main pumps, pressurizers, and steam generators. The surfaces of these devices emit radioactive radiation (dose rates ranging from 0.1 to 5 mSv/h), and the equipment layout is dense with complex piping networks. Previously, the inspection model relied on "specialized inspectors + lead protective gear," which presented three major pain points:

1. Inspectors had to wear lead suits (weighing up to 15 kg), limiting continuous operation to no more than 30 minutes per session, resulting in extremely low efficiency—a full inspection of a single main pump took three days.

2. Certain narrow gaps between equipment were inaccessible to personnel, leading to inspection blind spots. During the 2023 routine maintenance, corrosion at a pipeline weld was discovered that had not been detected in time.

3. Manual recording of inspection data was prone to errors, and data traceability was difficult.
   To enhance nuclear safety and inspection efficiency, the plant introduced two nuclear power plant–specific inspection robotic dogs in April 2024.

II. Implementation Process

1. Specialized Modifications and Deployment (April 5–April 20)

    The technical team carried out radiation-resistant modifications on the robotic dogs: the outer shell was constructed from lead–boron polyethylene composite shielding material, and core components were equipped with radiation shielding covers. Testing confirmed that the dogs could withstand radiation doses up to 10 mSv/h. They were equipped with a nuclear radiation detector, a high-precision industrial camera (capable of detecting weld defects as small as 0.1 mm), and an ultrasonic thickness gauge. Data connectivity was established with the plant's equipment management system. Using LiDAR, the reactor building was scanned to create a 3D digital map marking 120 key equipment units and 800 inspection points. Two inspection routes were predefined: a main equipment route and a pipeline system route. Inspection standards were set for parameters such as radiation dose, pipe wall thickness, and equipment vibration.

2. Trial Operation and Verification (April 21–May 10)

    The robotic dogs underwent trial operation during a unit outage for maintenance, focusing on inspecting main pumps, steam generators, and related pipelines. Two key functions were optimized during this phase:

• Difficulty navigating narrow gaps between equipment (minimum width 0.3 meters) was resolved by adjusting the folding angles of the robotic dogs' limbs, enabling flexible passage.

• The accuracy of ultrasonic thickness measurements, affected by oxide layers on pipe surfaces, was improved from ±0.1 mm to ±0.05 mm after adding an automatic grinding module.
  Verification by a third-party testing agency showed a 99.5% consistency between the robotic dogs' inspection data and manual inspection data.

3. Formal Application (May 11–Present)

    The robotic dogs operate both during normal unit operation and outage maintenance periods. During normal operation, they perform inspections of key equipment once every seven days. During outages, they assist personnel in comprehensive inspections. Inspection data is uploaded in real time to the equipment management system, which automatically generates inspection reports.

III. Application Results

1. Enhanced Nuclear Safety Assurance

• Inspection blind spot coverage was reduced from 15% to 0%. During the 2024 maintenance outage, two instances of pipeline wall thinning were detected (minimum wall thickness 21 mm vs. standard 25 mm), allowing timely replacement and preventing potential leakage risks.

• By replacing personnel in high-radiation areas, the robotic dogs reduced inspectors' radiation exposure by 90%.

2. Significant Improvement in Inspection Efficiency

• The time required for a comprehensive inspection of a single main pump was reduced from three days to eight hours, improving efficiency sevenfold.

• The duration of unit outage maintenance was shortened by five days, increasing power generation by approximately 150 million kWh per unit (valued at around 60 million RMB).

3. Reduction in Management Costs

• Four specialized inspector positions were eliminated, saving approximately 1.2 million RMB annually in labor and protective equipment costs.

• Digital storage of inspection data improved traceability efficiency by 100% and increased equipment fault diagnosis accuracy by 80%.

IV. Typical Scenario

    On June 15, 2024, during an inspection of the main pump in Unit 3, a robotic dog used its ultrasonic thickness gauge to detect that the wall thickness at a weld on the main pump outlet pipeline was only 20.8 mm, below the standard value of 25 mm. Concurrently, the nuclear radiation detector showed no abnormalities. The robotic dog immediately captured high-definition photos of the weld and uploaded the data. Based on the inspection report, the equipment management team developed a replacement plan. The pipeline section was replaced during the next unit outage. Disassembly analysis revealed that long-term erosion had caused corrosion at the weld. If undetected, this could have led to coolant leakage, potentially resulting in equipment losses worth hundreds of millions of RMB and serious nuclear safety risks. For this inspection, the robotic dog operated continuously for four hours without human intervention, and all data was automatically synchronized to the equipment records.