Projects
Projects: Projects for Investigator |
||
| Reference Number | NIA_NGGT0103 | |
| Title | Artificial Intelligence for Pipe Coating Inspection | |
| Status | Completed | |
| Energy Categories | Fossil Fuels: Oil Gas and Coal(Oil and Gas, Refining, transport and storage of oil and gas) 100%; | |
| Research Types | Applied Research and Development 100% | |
| Science and Technology Fields | PHYSICAL SCIENCES AND MATHEMATICS (Metallurgy and Materials) 50%; ENGINEERING AND TECHNOLOGY (Mechanical, Aeronautical and Manufacturing Engineering) 50%; |
|
| UKERC Cross Cutting Characterisation | Not Cross-cutting 100% | |
| Principal Investigator |
Project Contact No email address given National Grid Gas Transmission |
|
| Award Type | Network Innovation Allowance | |
| Funding Source | Ofgem | |
| Start Date | 01 November 2016 | |
| End Date | 01 November 2017 | |
| Duration | 12 months | |
| Total Grant Value | £142,000 | |
| Industrial Sectors | Technical Consultancy | |
| Region | London | |
| Programme | Network Innovation Allowance | |
| Investigators | Principal Investigator | Project Contact , National Grid Gas Transmission (100.000%) |
| Web Site | http://www.smarternetworks.org/project/NIA_NGGT0103 |
|
| Objectives | Improve quality and consistency of asset condition assessment data associated with the CM/4 process, enabling improved asset maintenance choices and investment decision making. The success criteria is based on the successful completion of the three deliverables detailed above: equipment classification, corrosion level classification and field testing to the business accepted standard. | |
| Abstract | Across the National Transmission System (NTS) there are approximately 450 sites, which must have the condition of plant coatings, painting and cladding inspected and assessed. Coatings systems provide the primary method of protection against corrosion on above ground assets. Inspection of coatings systems is required to ensure National Grid has visibility of asset condition and to ensure the risk associated with loss of containment is managed to a minimum. Inspections are currently carried out by technicians across the network this results in inconsistent categorization which leads to variable visibility of asset health condition. Inspections are carried out by qualified technicians at a frequency of anything between 1 month and 6 years depending upon the amount of risk associated with the coating system degradation and associated asset. As key information that supports investment decisions across the coating systems used on the NTS, there is an opportunity to improve the data collection, condition categorization accuracy, maintenance strategies and subsequent investment efficiency through application of artificial intelligence. Tractable holds patented machine learning technology that can be used to conduct visual inspections. The technology is applied in the automotive industry to assess insurance claims and predict optimal repair work. Visual inspection is common across the Gas Industry and within Gas Transmission one key process is specified in "The Specification for The Assessment and Reporting of Plant Coatings, Painting & Cladding Inspections for National Transmission System Assets" (T/SP/CM/4). This project will assess the feasibility of the application of Artificial Intelligence technology to the process defined in CM/4 and will be divided into three phases: Phase 1 - Equipment Type Classification: The current method of CM/4 inspection results in the collection of a high volume of digital photographs (approx. 30,000). This library will be used to train the algorithm to identify the seven different types of asset assessed by CM/4:General Paintwork and ancillary equipment Risers Flanges Pipe Supports Pit Wall transitions Cladding/insulation Ball Valve Sealant, body vent and drain lines The deliverable for this stage is the ability of the algorithm to categorize assets into the 7 sections detailed above to an 80% accuracy level. It should be noted that algorithm performance improves as the quantity of data provided increased and the learning increases. Phase 2 - Corrosion Level Categorization: The CM/4 process categorizes each of the seven asset types by risk associated to integrity. The algorithm will be provided with photographs and training from trained technicians and engineers to support enable it to categorize assets. The deliverable for phase two is the ability for the algorithm to carry out categorization of corrosion levels to an accuracy of 80%. Phase 3 - App Development, Field Testing and NG Acceptance: The focus of this phase is the transition of the algorithm from use in a development environment to a point of readiness for deployment to business as usual. This phase will deliver and App for use in iOS that provides an interface for users in the field to take photos, organize photos, conduct assessments and develop a report which is provided to engineers in an e-mail format. The current scope and funding will be evaluated at the end of phase 2 in order to assess whether the phase 3 scope, as currently scoped out, is appropriately scaled for field testing and business acceptance. The level of risk and uncertainty in developing the technique means it is possible additional time and funding will be required to achieve successful completion of phase 3.Note : Project Documents may be available via the ENA Smarter Networks Portal using the Website link above | |
| Publications | (none) |
|
| Final Report | (none) |
|
| Added to Database | 26/04/18 | |
