Predictive Build Modelling
Advanced spatial analysis and data models specializing in full fibre broadband network rollouts. Predict build types and costs enabling builds to adhere to budgets with high accuracy.
Innovative Planning Tools
Strategic planning tools, to automate high level plan creation including capacity and cost estimation which leads to streamlined planning and resource optimization.
Comprehensive Opportunity Analysis
Utilizing data models, market data and end-to-end reporting to create robust commercial business cases, enable effective management and tracking of opportunities.
Build Type Modelling. Predict Building Connection Types.
The purpose of the build type model is to accurately predict the connection types of premises to the Openreach network and enable precise build cost modeling. Using Ordnance Survey MasterMap data alongside Openreach PIA data, the model assesses each building's relation to network components like poles, span couplers, and chambers to predict the build or install type. It then generates a connection to the PIA network, with the predicted build type informing spatial cluster models, project cost estimates, and pre-planning tools. The model boasts a 95% positive prediction rate.
Detailed Cost Modelling. Driving build cost predictability.
The cost model is designed to calculate expected build cost for a variety of inputs, effectively supporting strategic planning and cost based decision-making of individual projects. It utilizes the Openreach PIA network data to accurately calculate the shortest possible meterage of network build required within each project, as well as the distribution route for each project. Cost estimates are produced, that can account for variations in build cost for MDU and blockages in PIA routes. This method ensures a detailed cost estimate is provided, which can then be overlaid onto various data inputs such as projects, clusters, or boundaries for comprehensive cost planning and decision-making.
Pre-Planning Tool. Rapid build prospecting
The pre-planning/prospecting tool is designed to refine data clusters into actionable packs, ready for planning partners to fully design. Its primary aim is to streamline the planning process, ensuring resources are focused only on premises that meet budgetary and strategic requirements, reducing wasted resources. It also facilitates a governance approval process for both planning and building phases monitoring costs throughout the process adjusting as required.
Pre-Planning can reduce time to assess 100k premises from >2 months to less than 7 days*
The tool operates on a QGIS-based input, processing all data centrally. Multiple users can define premises to be planned by drawing polygons on a map and including various optional data. Based on these inputs, it calculates and plans the job, outputting results to the database and viewed back QGIS layers. The tool's adaptability allows for varying approaches depending on user input. For instance, it can select the nearest cabinet for planning if one isn't specified or link packs in a daisy chain for efficient resource distribution. It also accounts for the presence of splitters within packs, planning connections and resources accordingly. This tool significantly contributes to strategic planning, allowing for efficient and targeted deployment of build opportunities.
*Based on creating low level designs on 100% of premises then reviewing completed plans
Spatial Clustering. Strategic Opportunity Identification.
This innovative approach to build opportunity identification leverages the precision of Ordnance Survey building data to intelligently group buildings into viable build opportunities. Utilizing spatial capabilities of PostgreSQL and PostGIS, buildings are systematically clustered and costed, drastically reducing the time to identify viable opportunities, streamlining the planning process by massively reducing wasted planning resources.
Video demonstration of spatial clustering running on ~4 million premises in London.
Skills used
- PostgreSQL
- Database design & implementation.
- Bespoke functions & procedures
- PostGIS
- Advanced spatial analysis
- Spatial data creation
- Python
- Data transformation and manipulation
