The future of linear infrastructure is a human + machine team

Discover how Mott MacDonald and Continuum Industries combine human and artificial intelligence to save 40% on design time for water pipelines and reduce the whole life cycle costs of multi-million dollar projects by more than 10% compared with traditional methods.

By Adam Cullum (Mott MacDonald) & Matt Blythe (Continuum Industries)

Plenty has been written about how artificial intelligence (AI) will transform infrastructure design and delivery, but what the industry really wants to know is how large the return on investment in AI is and how soon the returns can be realised.

At Mott MacDonald, we devised a test case for our engineers to solve using both traditional methods and AI, to answer two key questions that drive return on investment.

1) What are the quantifiable savings from using AI to design infrastructure schemes?  
2) Can the same AI design all types of infrastructure in different locations?

1. TEST CASE PROBLEM
2. APPROACH 1: WITHOUT AI
3. APPROACH 2: WITH AI
4. RESULTS
5. CONCLUSIONS

Test case problem

We set a group of engineers in our water team a classic early-stage design task:

‘Given multiple start and end points for a pressurised water pipeline, find the best possible horizontal and vertical route alignments for each that result in the lowest whole life cycle costs (combined construction and operational costs). You may use as much GIS data as you consider necessary and your designs must comply with engineering requirements such as minimum depth of cover and gradient restrictions.’

In other words, we gave them a blank piece of paper and asked them to come up with outline designs for a water pipeline scheme from scratch. Let’s see how they did.

Example horizontal route alignments prepared with (yellow) and without AI (blue)

Results

To compare the results from both approaches, Continuum Industries used their AI algorithms to evaluate the whole life cycle costs of all of the horizontal and vertical route alignments using the same design and costing rules and specified parameters for each.
In all three cases, the best designs produced through the combination of our engineers’ expertise and Continuum Industries’ AI algorithms resulted in significant savings in the whole life cycle costs of each pipeline compared with traditional methods. The lengths of the pipelines ranged from around 15km to 35km.

Estimate Savings in Whole Life Costs using AI
Pipeline 1 - 10.7%
Pipeline 2 - 8.5%
Pipeline 3 - 2.8%

Download the full case study at the top of this page to understand why there were such significant whole life cycle savings.

Example vertical profile for route alignment

Download
the Full Case study
Learn more about Moata Route Optimiser »

Any questions?

Let us know if you are considering using Optioneer for your project and our team will be happy to advise how it would best fit into your workflow or provide a quote.

Contact us