Data can now be collected and mined by artificial intelligence to enhance predictability in the infrastructure planning process. InEight’s Chief Design Officer Dr. Dan Patterson explains how companies can make the most of it.
In 2017, The Economist reported the world’s most valuable resource is no longer oil, but data.
The global influence of multinational tech giants has meant information is becoming a higher-valued commodity, and its influence is even greater with aggregation.
When data is organised cohesively, it can help inform a business’s investment decisions, and assist it in staying ahead of the competition and growing its market share.
In today’s world, quantitative data is often used to aid predictability, so that companies can advertise products to the right people at the right time. By generating more data, a firm has more scope to improve and target its products, which attracts additional users and generates even more information.
This same principal can now be applied to construction. In recent years, the industry has been exploring the ways in which data can enhance project performance.
One of the companies at the forefront of this movement is InEight, a capital project management solution provider that helps construction companies unlock data capabilities and enhance predictability in project planning.
InEight Chief Design Officer Dr. Dan Patterson says being able to better predict the future relies heavily on capturing historical information. By learning from past data, a project can be planned and executed with greater accuracy, enhancing the outcomes of similar projects in the future.
“With the advent of unstructured data storage, we can now store and, more importantly, mine much richer and more varied types of historical data, and that gives us more confidence in our forward-looking predictions,” Dr. Patterson says.
“The more data you have, the more accurately you can predict the future. It’s OK to have good historical data, but it’s also wise to have information on projects that perhaps didn’t go so well, because those projects were likely executed to reality.”
Predictive planning is the ability to leverage patterns from the past to more accurately forecast the future.
“The more accurate our project forecast, the more alignment and harmony there is between owners, contractors and stakeholders,” Dr. Patterson says.
Traditional planning and scheduling tools are aimed at generating the best-case scenario outcome – part of the reason he believes projects are often delivered late. It’s not so much a matter of poor execution, but rather, the fault of overly optimistic and ambitious plans.
“Most projects carry what we call a risk register which, in its simplest form, is a spreadsheet-based list of things that could go wrong.
“With this, the computer can identify poorly executed projects and the risks associated with those, and apply them as a warning for a similar new project,” Dr. Patterson says.
He says that with the likes of artificial intelligence (AI), computers are now much better at pattern recognition, increasing predictive capabilities.
AI serves as a vital tool for achieving project plan predictability, as it provides an engine that can intelligently sort through huge volumes of historical data and start to make inferences from it.
“In the real world, the computer has the ability to critique a project schedule as it is being built. If the computer has access to historical data and can understand context, it can benchmark and critique the plan and make suggestions for improvement,” Dr. Patterson says.
Its power relies on an embedded AI engine that constantly guides plans through an iteratively improving forecasting cycle. A range of different files can be uploaded to the system from previous projects for analysis.
The system also provides a collaborative platform for team members to provide their expert opinions.
“We call this human intelligence, and further, our tool now features a risk analysis system that offers suggestions as to which risks and opportunities a project should track to help identify potential unknown factors,” Dr. Patterson says.
“With risk intelligence, we are able to generate a risk-adjusted forecast, resulting in a much more realistic model and predictive outcome of the future,” he says.
The synergy of artificial, human and risk intelligence in InEight’s platform results in what’s termed as “project intelligence,” which combines these elements to predict the most likely outcome.
“It would be wonderful to think that in the future, as part of the bidding process, not only do you present your bid, but in addition, you have to show how you got there and how realistic it is compared to your historical data.”
He says if companies get the forecast wrong, it sets up the project for failure even before day one of execution.
“Organisations have huge amounts of expertise and experience, but very few today recognise the importance of digitalising this for consumption by modern AI-driven tools that then help drive predictability.”
Not only is planning important before a project begins, but it becomes integral during on-site execution, as there are still questions around how long something will take, or how many resources and crews are needed.
Dr. Patterson notes that InEight likes to view artificial intelligence as augmented intelligence, as the technology is not designed to replace planners, but rather, enhance their decisions and help them perform their jobs more effectively.
“Artificial intelligence is simply empowering humans to do their jobs more efficiently and make more accurate predictions,” he says.
“No matter how close to true human emulation a computer can get, humans will own and have ultimate responsibility for projects.”
As that message gets out, scepticism around the notion of AI’s use in road construction is beginning to dissolve.
“I honestly think two years from now you will be an exception if you are not using some form of artificial intelligence to review prior performance and help predict the future,” Dr. Patterson says.
To learn more about InEight’s planning, scheduling and risk solutions, visit ineight.com.
Related stories:
- Connecting systems and site
- Austroads releases revised Data Standard for road asset managers
- Austroads harmonises road asset data