Column 10 of the "15th Five Year Plan for Urban Renewal" clearly proposes the improvement of the safety guarantee capability of urban gas facilities, requiring the strengthening of intelligent supervision throughout the entire chain, promoting the large-scale and professional development of urban gas enterprises, accelerating the digital, networked, and intelligent transformation of municipal infrastructure, and consolidating the foundation of urban safe operation. At present, various regions are synchronously implementing the construction of the "six networks" in cities, accelerating the digital upgrading of urban underground pipelines, energy security networks, and urban safety operation systems, and promoting the transformation of urban safety governance towards intelligence, refinement, and full process supervision.
As an important component of the urban energy security system, bottled liquefied gas is widely used in catering operations, residents' lives, and other scenarios. Its safety management is directly related to the safety of people's lives and property and the stable operation of the city. At the same time, policies such as the "Guiding Opinions on Strengthening the Safety Management of Bottled Liquefied Petroleum Gas" and the "National Urban Gas Safety Special Rectification Work Plan" have been successively introduced, which have put forward higher requirements for risk prevention and control, full process traceability, and intelligent supervision in the liquefied gas industry.
The bottled liquefied gas business runs through multiple stages such as storage and distribution, filling, distribution, security inspection, use, and recycling. Accidents such as the gas explosion at Wuxi's "10-13" snack shop and the gas leak at Yinchuan's "6.21" barbecue shop have deeply exposed the stubborn problem of "small, scattered, and chaotic" governance in the bottled liquefied gas industry. The entire industry chain has multiple circulation links and wide regulatory coverage, coupled with limited regulatory power and frequent violations and chaos, resulting in a large stock of various hidden dangers and high accident risks. Small problems can easily evolve into malignant safety incidents. Among them, there are many violations, poor quality of security checks, and difficult closed-loop rectification of hidden dangers, which are prominent challenges that gas safety supervision has long faced. The traditional manual supervision method is no longer able to meet the current demand for refined security governance. Focusing on prominent challenges in the industry,Chen'an Technology uses core technologies such as AI visual recognition, data intelligent analysis, and full process traceability to build an AI intelligent supervision system for bottled liquefied gas, promoting the industry's upgrade from "manual inspection" to "intelligent warning and proactive prevention and control".
AI Empowerment: From 'People Finding Problems' to' Systems Proactively Discovering '
Intelligent recognition of AI violations
The liquefied gas industry has long been plagued by problems such as unlicensed distribution, illegal filling, cross regional distribution, and vehicle overloading. Traditional manual spot checks have limited efficiency and many omissions, making it difficult to detect violations in a timely manner.
Build a series of AI violation recognition algorithms on the platform to achieve 7 × 24 hour intelligent recognition of 58 types of abnormal behaviors in 6 major business processes including filling, delivery, and security checks. At present, the accuracy of AI recognition exceeds 95%, and it can automatically identify more than 400 abnormal problems per day, effectively improving regulatory efficiency and risk detection capabilities. Some cities have carried out precise law enforcement and special rectification based on AI analysis results, conducted scheduling interviews with more than 10 related enterprises, punished more than 100 people, and promoted the standardized operation of the industry.
AI security check photo intelligent quality inspection
The daily number of security check photos is huge, and the quality of photos varies greatly. There are problems such as blurry security check photos, repeated use of the same photo, and inconsistency between security check photos and inspection items. Traditional manual quality inspection has low efficiency, high omission rate, and difficulty in ensuring authenticity.
The platform is based on an AI photo recognition model to achieve fully automated quality inspection of security check photos, using AI intelligent quality inspection to solve the problem of massive security check photo supervision. The overall accuracy rate of security check photo quality inspection exceeds 93%. At present, about 1 million security photos can be automatically inspected every month, and about 140000 non-standard photos can be identified. The system can complete the audit workload of about 500 people per day in just one hour, greatly improving the authenticity and standardization level of security checks. At the same time, the platform can automatically generate quality analysis reports for enterprise and employee security checks, assisting in precise supervision and punishment of more than 100 people in multiple regions, accurately identifying violation clues, and providing data support for regulatory enforcement and punishment.
AI hidden danger rectification intelligent review
Whether the rectification is truly implemented has always been a key and difficult point in the management of liquefied gas safety hazards. For a long time, there have been common problems in the process of rectifying hidden dangers, such as blurry photos, inconsistent scenes before and after rectification, and repeated uploading of photos. It is difficult to detect phenomena such as "paper rectification" and "fake rectification" in a timely manner.
In response to the above issues, the platform relies on AI intelligent comparison and
Image recognition technologyAutomated verification and analysis of photos before and after hidden danger rectification can accurately identify abnormal situations such as blurry photos, mismatched scenes, repeated use, and false rectification, with a review accuracy rate of over 90%. At present, the platform can automatically review more than 8000 pieces of hidden danger information per month, using digital and intelligent means to promote the "real implementation and real closed loop" of hidden danger rectification, further improving the effectiveness of liquefied gas safety governance. Some cities have discovered over 600 abnormal rectification issues through the platform, and have simultaneously carried out hidden danger supervision and closed-loop rectification to further enhance the effectiveness of hidden danger governance.
Implementation of digital intelligence and upgrading of urban safety governance capabilities
As of the first quarter of 2026, the platform has been implemented and applied in over 20 cities across the country, gradually building a liquefied gas intelligent supervision system covering the four levels of "province city district county" linkage. Currently achieved:
Comprehensive access to basic data
More than 9200 active workers per day
The number of violations has decreased by about 50%
The non-standard rate of security check photos has decreased to 22%
The false rectification rate of hidden dangers has decreased to 20%
At the same time, the intelligent supervision mode has received special attention from CCTV's "Economic Information Broadcast" and attracted multiple regions to carry out learning exchanges and experience reference.
From "manual inspection" to "intelligent supervision", from "post disposal" to "risk prevention", Chen'an Technology empowers supervision with AI, safeguards safety with technology, and continuously reshapes the gas safety governance model with three major AI capabilities. In the future, Chen'an Technology will continue to focus on the requirements of urban safety construction during the 15th Five Year Plan period, deepen the integration of AI algorithms and gas supervision scenarios, promote the digital upgrading of urban lifeline safety systems, and use technological strength to build a strong urban safety barrier and protect the "fireworks" of every household.
