Existing design approaches to linear assets are geared towards quickly identifying a feasible option early in the design phase and then uncovering problems and proposing solutions to them during more detailed design. However, as linear assets are increasing in their complexity and end-clients want more certainty early in the design phase to lower construction risk, the status quo is clearly not sustainable.
Optioneer offers a new approach to desktop optioneering studies for linear infrastructure. It lets project teams automate the existing optioneering process, efficiently capture a high level of geospatial and engineering detail and iterate through thousands of design options to find the best ones.
For each project, the first step is to select a set of parametric design rules and unit costs that will be used to design the linear asset. Once users configure their preferred parameters for each design rule and unit cost, their ‘model’ is ready to use.
At this point, users can save their model and move to the next step or alternatively, build and save more models with different sets of parameters configured or entirely different combinations of parametric design rules and unit costs.
Users can import their geospatial data into Optioneer to capture key geographic features, topography, existing assets and other relevant information about the project area.
Once this is complete, users can link individual geospatial data layers to their models to configure penalties or other measures and designate individual geospatial data layers as ‘no-go zones’ to be avoided.
To start generating design options, users define the start / end points of their linear asset and any waypoints within their corridor or alignment and select the model they wish to use to drive the optimisation process.
Typically, it will take 1-3 hours to return multiple design options to the user. However, users do not need to remain signed in while the optimisation completes and multiple optimisation runs using different models and combinations of start / end / waypoints can be run at the same time. Users can also check the progress of all of their optimisation runs in the home screen for their project.
The optioneering engine is non-deterministic optimisation process and works by rapidly generating, evaluating and evolving millions of possible design options using the detail provided to return multiple good design solutions.
Once the optimisation runs are complete, users can evaluate and compare the multiple design options generated. Each option is displayed in the interactive map interface alongside the imported geospatial data as illustrated in Figure 6 below, which shows an example of an underground power cable project in Optioneer.
Visualisations of key performance indicators (e.g. construction cost, number of crossings, excavation volumes) based on users’ models are provided for each design option. Users can also analyse specific points along the chainage of the corridor or alignment on both the horizontal and vertical profile in more detail. Furthermore, if users want to refine some individual design options, they can draw (or import) new options directly within the map interface and submit for rapid evaluation.
Finally, once users have selected their preferred design options to take forward to more detailed design, they can export these options and the associated metadata from Optioneer into GIS or CAD software for further analysis and refinement. Export file formats are flexible and can be configured to suit users’ needs.
Optioneer integrates constraints and models from all disciplines to give a complete view of the project.
Each project has multiple criteria and stakeholders that have to be taken into account. Optioneer gives you the data needed to make informed decisions.
Define parametric models of your own optioneering process and re-use them on multiple projects to save time and money.
Mott MacDonald and Continuum Industries have teamed up to combine the best of human and artificial intelligence (AI).
This case study showcases how engineers can apply AI to do much faster, more detailed optioneering at the start of any linear project and save millions in project costs.