Transforming sewer management with advanced geospatial and analytical tools
At a glance
GHD Digital streamlined the City of Bishopville’s sewer architecture data collection using geospatial insight tools that enabled more efficient management and analysis of their infrastructure. Incorporating analytical tools increased data accessibility, enhanced accuracy and provided better planning and asset management.
The challenge
The historic City of Bishopville, located in South Carolina faced a few challenges with variability in the quality of scanned drawings and the complexity of recognizing and classifying diverse sewer features.
They had an ambitious goal to convert all hand drawings of their sewer systems to shape files that were compatible with geospatial tools, like ArcGIS in less than two months to improve their management and analysis.
Another challenge they faced was with the accuracy and accessibility of their data for city planners and engineers. This also required better integration of their existing data into ArcGIS systems for enhanced spatial analysis and decision-making.
Our response
GHD Digital’s digital intelligence team provided the City of Bishopville with a straight forward and repeatable approach to better understand and improve their sewer systems. Using optical character recognition (OCR), we extracted text information from scanned PDF drawings. Artificial Intelligence (AI) automated the recognition and classification of sewer features, while computer vision identified and interpreted graphical elements in the drawings.
Our methodology for data collection began with scanning all hand drawn sewer drawings into a PDF format. Then we enhanced the scanned images to create more OCR accuracy. By reducing noise, contrast adjustment and alignment correction, we were able to leverage OCR technology to extract textual information such as labels, dimensions and annotations from the hand drawings. Using advanced image preprocessing techniques, we improved the quality of scans before applying OCR and computer vision.
AI and computer vision integration were developed to recognise and classify various sewer features, for example, pipes and manholes from the drawings. We then applied computer vision techniques to interpret and convert graphical elements into vector data (i.e. the digital format used to represent the shapes and locations of graphical elements within the sewer maps). AI models were then trained on a diverse dataset to improve recognition accuracy and employed iterative testing and refinement.
Once the files were created, we streamlined the extraction and interpretation of data into shape files that were compatible with ArcGIS, ensuring accuracy through validation checks. Transforming the files to ArcGIS systems enabled spatial analysis and visualisation with ease.
The impact
Timely and accurate data collection and analysis are vital to projects and programs planned and implemented by utilities. The City of Bishopville now has a fully digitised sewer system with improved data accessibility and accuracy, enhancing the city’s ability to manage and analyse its sewer infrastructure effectively. This has provided a better understanding of spatial deliverables, making project details more digestible for city planners and engineers through ArcGIS.
The integration of ArcGIS allowed for more efficient management and analysis of sewer infrastructure, aiding in maintenance and planning efforts. Leveraging advanced analytical tools significantly reduced manual errors, resulting in more accurate shape files to improve data-driven decisions.
Connect with us to learn more about leveraging geospatial insights to streamline decision-making and improve city planning.