With construction projects growing in size and complexity, traditional risk assessment tools often fail to accurately predict uncertainties, leading to budget overruns. InEight expert Nate St. John explains why a collaborative, risk-adjusted approach assisted by artificial intelligence is the answer. Tara Hamid writes.
The dystopian Hollywood movies depicting a future led by artificial intelligence (AI) and machine learning often portray a world where robots have run amok, overtaking humans [with their limited intelligence] in running the affairs of the world.
While the fear might not be entirely unfounded given the dilemma of our time around use of social media and its ever-growing data pool, one reality cannot be undermined: Artificial intelligence is already here and it’s helping humans take better decisions in their everyday lives.
As Nate St. John, Director of Planning, Scheduling and Risk Products at InEight observes, it would be a loss not to use that unbiased machine intelligence to assess risks and bring down the level of uncertainties on construction projects, particularly the larger, more complex ones.
Today’s capital construction projects have grown bigger and more complex from just a decade ago. Multibillion dollar transport infrastructure projects are no longer an exception, but a norm. With that increase in size of projects come bigger uncertainties and risks of budget overruns.
In its 2020 report, ‘The Rise of Megaprojects: counting the costs,’ Grattan Institute announced the arrival of an era of megaprojects in Australia, where two thirds of the government’s transport infrastructure projects are now valued over $5 billion. The report found an overrun of $24 billion on just six mega transport projects [within the report’s study period], where bigger projects had a higher risk of outrunning budget estimates.
Risk assessment on megaprojects
Having had involvements with the planning and scheduling corporate steering committee at one of the largest construction and engineering firms in North America for several years prior to joining InEight, St. John has had first-hand experience with pre-planning, baseline development, progress and execution.
He says the biggest risk of relying on the traditional project management tools to assess risk in mega construction projects is that they fail to incorporate active risk adjustments through the project lifecycle.
“In the traditional planning approaches, such as a critical path method, there’s a heavy focus on planning from left to right, where you start from one point and draw the path of construction. The problem with relying solely on that approach is that it subjects you to areas of compounding delays, where one incident can have a ripple-on effect down the line,” he explains.
Having disconnected risk registers and schedules is another critical factor that can lead to knowledge gaps, resulting in overly optimistic estimations, St. John observes.
“Traditionally, the process of establishing a risk register and then mapping that to the schedule has always been a separate exercise. That’s because they’ve always used separate tools. The gap between a disconnected risk register and the schedule is usually filled by experienced team members accounting for risk manually or applying their ‘gut feel’ against contingency items. As seasoned as your project team may be, there will be areas missing coverage on risk exposure, resulting in unnecessary risk consumption.”
The other area where projects can get into trouble, he observes, is where the schedulers retain most of the responsibility with themselves, without establishing an element of active risk management in collaboration with the project team.
“I’m a firm believer that communication and planning have to scale with project’s size and complexity. And so, with capital projects getting bigger and bigger, if you don’t put the upfront effort into establishing an honest communication and planning culture within your team, you are always going to lag behind,” he says.
Filling the risk knowledge gap
So how do modern risk management tools bridge the risk knowledge gap to minimise uncertainty on projects?
As St. John explains, a modern planning tool such as InEight’s integrated project planning, scheduling and risk analysis software can combine AI, machine learning (ML) and a collaborative approach to create a risk-adjusted plan to achieve more certainty in project execution.
“InEight Schedule, which is InEight’s planning, scheduling and risk analysis software, is designed around the concept of risk-adjusted forecasting to help improve project predictability and to provide a ‘most likely’ scenario rather than a ‘best case’ scenario.”
In this risk-adjusted approach, St. John explains, the outcome might be more conservative, but it will also have a higher degree of certainty.
“For a megaproject, you need to be able to say with at least 75 to 80 per cent confidence that the predicted outcome is going to happen. That’s what InEight’s tool offers, using a consensus model based on feedback from experienced team members, as well as relying on AI and ML for additional insights,” he says.
He then goes on to explain each of those capabilities further.
“One of the things that we introduced at InEight Schedule is the ability to solicit feedback from experts anywhere in the world. If you envision the phase of the project when you have an established baseline schedule, the scheduler can then send a digital invite to several planning experts anywhere in the world, asking them to add their markups. In this way, you can capture a varied consensus pool without affecting your underlying schedule and take in all of those opinions into factor,” he explains.
“The good thing about this approach is that people can take their time, away from peer pressure, to provide their honest opinion. From my experience of facilitating numerous risk workshops for construction companies, when you take a poll in real time, the loudest voice in the room often wins. Our mission is to democratise access to risk assessment so that where input used to be reserved for expensive consultants, it can now be extended to a bigger crowd to create a knowledge library.”
Computers to the rescue
The other unique ability in InEight Schedule is the incorporation of AI and ML to assist humans with decision-making. Here, St. John says, the computer’s knowledge can augment the scheduler’s experience to better recognise potential risks.
“The AI and the ML components of the tool means that the software can use data from your company’s past experiences to prompt the scheduler to consider risk factors the company has previously experienced on similar projects. It is, of course, up to the planner to accept the software’s suggestion or not. The end user always has the ability to say: no, this isn’t appropriate for my scenario,” he explains.
“The ML component of the tool also means that when you reject a suggestion, the computer will go back and re-weigh the algorithm so that the next time that scenario happens, it tends to be smarter and provide more meaningful suggestions. In this way, the more data-rich companies become, or in other words, the more they place clean, good data into the system, the better insights and outcomes they can get.”
For those who may still be sceptical about the power of AI, St. John reiterates that computers are more un-biased than humans.
“As human beings, there are cognitive biases in all of our decision-makings. Computers are not biased, they are using data, their algorithms can be flexible, so they get smarter overtime. But if you need somebody, a voice of reason, sometimes letting technology play the devil’s advocate and show you a different option just sets you up for more insight, which sets you up to make smarter business decisions.”
Scheduling for the future
While there isn’t a catch-all solution to mitigating all risks on projects, St. John believes relying on old techniques and hoping they would be enough to provide a realistic picture of today’s complex projects would be like wearing ‘rose-coloured glasses.’
Taking a risk-adjusted approach to planning and scheduling, he says, is even more critical given the changing anatomy of construction contracts.
“Not a long time ago, the construction sector was inclined to use a design-bid-build model of contracts – which is what some people also referred to as rip-and-read contracts. These days and with broader use of third-party agreements, the risk has been shifted mostly to the contractors to manage engineering, procurement and construction,” he says.
“While project teams normally do a decent job of generating a pre-bid risk matrix, the critical element of active risk management can often become a mere afterthought. This is what we refer to in the industry as wearing ‘rose-coloured glasses’ and hoping everything is going to be fine. That’s where things often go to the sideways.”
With these increased contractual risks, taking a risk-adjusted approach to planning would only make sense, St. John concludes.
“Today’s collaborative tools allow you to democratise access to technology so that you can establish a self-perpetuating risk assessment culture within your organisation and rely on it as a baseline to mitigate risks for your projects. Particularly in a sector like construction that’s so iterative and suited to scenario-based modelling, why would you not leverage that additional insight?”
This article was originally published in the February edition of our magazine. To read the magazine, click here.
Related stories:
- InEight’s innovations facilitate construction 4.0 capital projects
- InEight: The power of connected data
- InEight releases platform enhancements for deeper project insights