Guimu road disease robot detection system
With the rapid development of artificial intelligence, robotics, nondestructive testing sensors, 5g, big data, Internet of things, cloud storage and processing technology, highway management and maintenance is showing from artificial semi-automatic sampling detection to intelligent comprehensive and accurate detection, from low-frequency static evaluation to high-frequency dynamic evaluation and early warning, from single information analysis to multi heterogeneous data fusion evaluation, from experience based evaluation From passive extensive maintenance to accurate maintenance based on data prevention, from phased discrete maintenance to life cycle maintenance.
Under this development background, Shanghai Guimu robot Co., Ltd. (hereinafter referred to as “Guimu”) officially launched road disease detection robot products in 2018. With the cutting-edge technology of “robot + nondestructive testing + artificial intelligence + 5g + big data”, adhering to the idea of “Ai assisted intelligent maintenance”, we strive to build disease level accurate detection intelligent equipment for the transportation industry, fill in the gaps in “special level investigation, disease level detection, precision level diagnosis, big data level management” of the industry, and jointly build a “prevention” platform with the upstream and downstream of the industry It is the solution under the scenarios of “comprehensive maintenance”, “repair maintenance” and “road life cycle management”.
Robot detection system for road diseases — CT machine for road structural diseases
Guimu road disease robot detection system is equipped with a variety of non-destructive detection means to complete the data collection of pavement and subgrade diseases, especially structural diseases with full autonomy, full coverage and high precision.
The high-performance robot platform which can eliminate the interference of human factors is equipped with array three-dimensional ground penetrating radar to realize the detection of structural diseases within 3 meters underground. The reliable positioning of centimeter level multi-source data fusion is adopted to provide intuitive and accurate disease location information for design and construction units. Combined with the sub millimeter level apparent disease data collected by high-definition industrial camera, the system can clearly present the apparent and underground structure status of any location of the road, and describe the type, size, severity and location of each disease in detail Finally, the high-precision detection of 100% full coverage of a road is realized.
On the whole, in the aspect of data acquisition, the system presents three advantages.
High data accuracy: sub millimeter level apparent sampling accuracy, centimeter level underground structure disease sampling accuracy, combined with centimeter level multi-source data fusion and reliable positioning, ensure the high accuracy of the system data as a whole.
High data density – especially in the road direction, the detection of GPR on the road is high-frequency, uniform, continuous, carpet type full coverage, so the data density can reach the exponential level of traditional semi-automatic detection equipment, which provides the premise for the accurate analysis and diagnosis of back-end artificial intelligence analysis algorithm.
High consistency of data – in this way, under the periodic detection, it is possible to compare and analyze the deterioration trend of diseases at the same location in the time dimension. At the same time, good data accuracy and consistency make the system as a high-quality data entry, which is very suitable for data access to traffic maintenance big data system or maintenance cloud platform.
“Appearance + structure” 3D data model
Three dimensional real modeling of road
“Accurate modeling + intelligent diagnosis” — make the whole road disease clear at a glance
On the basis of acquiring the big data of the above diseases, Guimu’s self-developed artificial intelligence algorithm can automatically process the massive disease data, automatically model the three-dimensional real scene of a single frame, single point, single apparent picture and GPR map, and integrate the accurate location coordinate information, and then generate the real scene panorama of the whole highway disease distribution, which makes the disease information more accurate Intuitive and comprehensive. In particular, it is easy to grasp the transverse cracks across multiple lanes or large area of void. On this basis, the artificial intelligence algorithm will identify and diagnose each disease independently, and its recognition accuracy is at the leading level in the industry.
“Appearance + structure, part + whole” — multi dimensional fusion analysis
Guimu detection robot products realized the simultaneous detection of “apparent disease + hidden disease” in China in early years, and complete the accurate fusion analysis and superposition presentation through the algorithm, which brings a more three-dimensional analysis perspective of road health status for highway maintenance, and helps the management and maintenance units to explore the correlation and evolution law of disease from the outside to the inside or from the inside out.
From the local point of view, the system can analyze the apparent disease and the structural disease in the same position below it, especially suitable for investigating the root cause of the recurrence of the apparent disease. For example, whether a crack in a surface layer is a simple apparent disease or a reflection crack caused by a structural problem below, different treatment schemes are needed for different situations.
From the overall point of view, a single disease can be comprehensively analyzed with the surrounding other diseases and the overall situation of the road, which is more conducive to finding the real lesions and treating both the symptoms and the root causes.
Splicing analysis results of local diseases on a national and provincial trunk road (two apparent cracks above and underground hidden diseases in orange block)
Overall disease analysis chart
Exploring “Guizhou model”
In terms of the application of four new technologies of “robot + Intelligent Maintenance”, Guizhou Province, as a pilot province of traffic power, adheres to the open mind of taking the lead in scientific and technological innovation and takes the lead in the field of national highway intelligent management and maintenance. Zunyi, Anshun, Bijie and other places actively promote and realize the landing application of road disease robot detection system. Through robot means, they have successively completed the tasks of disease fine investigation of key road sections, special investigation of maintenance, periodic detection of preventive maintenance, inspection of road safety hazards, layer by layer evaluation of new road structure, digital modeling of road underground pipelines, etc., with preliminary accumulation It also lays a solid foundation for the realization of digital management and centralized management and control of road life cycle in the whole province in the future, and actively promotes the formulation and implementation of relevant standards.
In the future, Guizhou Provincial Highway Bureau will also explore the application and upgrading of robot precision detection in the whole province in combination with the pilot construction of a transportation power, and further explore a new data-based highway management mode based on the construction of robot Precision Detection Demonstration Area in China.
Photo group: in 2019 alone, Guizhou Province has completed the inspection of tianjiawan tunnel of Zunyi Highway Bureau, Meitan Yuqing section of national highway 243 and Zhenning section of provincial highway 314 of Anshun Highway Bureau.
Using “precision” to serve repair and preventive maintenance
Up to now, Guimu road disease robot detection system has been applied in nearly 30 highway projects in 10 provinces of China, including Sichuan, Guizhou, Zhejiang, Shanghai and Beijing, covering expressways, national and provincial trunk roads and rural roads. As an emerging technology under the policy of promoting the four new technologies of highway, Guimu has cooperated closely with highway management, testing, design, construction, maintenance, scientific research institutes, societies and other units at all levels in technical research, application mode, business upgrading, emerging management and maintenance mechanism, standards and specifications, and made substantial progress.
Disease distribution map based on robot detection results
In the field of repair and maintenance, Guimu cooperates with the industry to build an overall solution of “road network level general survey + precision detection + precision design + precision construction + effect retest + data management”. Through the maintenance logic of “whether there is a problem – what is the problem – where is the problem – what causes the problem – how to repair the problem – whether the problem has been really solved” with data, it can assist the large and medium-sized maintenance of roads, improve the quality and efficiency, and improve the utilization efficiency of maintenance funds.
In the field of preventive maintenance, we advocate “preventive treatment, preventive treatment and general accounting”, and use accurate data to guide early prevention and control, so as to reduce the waste of maintenance resources caused by “treatment after outbreak”.
At the same time, based on the big data of road health, a backtracking mechanism guided by the development of road diseases is initially established to analyze the law of performance degradation and the causes of diseases. Cooperate with the establishment of Provincial Highway Disease big data archives, scientifically formulate maintenance investment plan, and carry out maintenance effect analysis.
Guimu fully adheres to the technical route of “Ai +”, creates high-quality highway health big data entry, fully maintains the openness of the system, and combines with highway maintenance cloud technology to realize road health trend research and judgment and road asset life cycle management. Guimu will be guided by the highway development plan of the 14th five year plan, serve the “national evaluation” as the goal, and contribute to the technological innovation and development of the national highway industry with intelligent products and services.