Contractor uses AI to capture knowledge of retiring engineers | New Civil Engineer

2023-02-05 17:24:01 By : Ms. Anna Chen

With projects depending heavily on the knowledge of specialists, a Hawaiian contractor has found a way to capture and store that knowledge before they retire.

E xperienced estimators could look through a project’s plans and tell you in a few hours the rough cost and duration of a project,” says Hawaiian Dredging Construction Company (HDCC) building information modelling (BIM) manager Chris Baze. “But doing that in their head, based on experience, means that it [their thinking process] is not documented. And if they retire, all the knowledge goes out of the door with them.”

The long learning curve associated with scheduling construction projects means that knowledge lost by someone retiring is not easily replaceable by new recruits.

“You probably have to build about 10 high rise buildings before you can really say that you know what’s going on and how all the pieces work together,” Baze points out.

For this reason, HDCC looked into ways to capture and store that knowledge. The firm found that it could achieve that using Alice, a platform developed by United States-based technology provider Alice Technologies.

W e’re able to take all of the lessons learned, rules of thumb and actual operations knowledge and document them in the form of a recipe

Alice uses artificial intelligence to create thousands of project schedule options and generate “what if” analyses. Contractors can use these to test variations in the way they can construct a building or piece of infrastructure.

HDCC used the platform to schedule the construction of two high rise residential developments in Honolulu, Hawaii, exploring the “what if” scenarios generated.

The way it is using Alice to digitise the scheduling knowledge of its experienced employees is through the creation of what Alice Technologies calls “recipes”.

When BIM models are imported into the platform, users can select a specific core project element – for example concrete column construction – to create a dedicated recipe. The recipes define all of the different tasks required to build that element and can add information about the materials, equipment and labour needed for each task. Tasks are presented in a network diagram illustrating the sequence to be followed.

HDCC hosts sessions where its experienced superintendents and subcontractor teams members contribute to the creation of the recipes. These are known as pull planning sessions, a collaborative lean production practice through which a team starts with the end goal and works backwards to analyse construction methods and identify potential issues. The aim is to distill experience into repeatable formulae.

Baze says that the sessions look into different construction scenarios for particular elements and examine all of the factors that can impact them.

“We say ‘here is a column and here is the sequence of construction. If that column is twice as tall, you will have to introduce additional components of scaffolding. Even though that column is only twice as big, it might need three times more time to complete it.’

I t does not only educate the new workers, it gets them productive faster

“We’re able to take all of the lessons learned, rules of thumb and actual operations knowledge and document them in the form of a recipe.”

The input from these sessions is then added directly to Alice.  

Alice Technologies chief marketing officer Phil Carpenter is impressed by HDCC’s decision to use the recipes to capture the knowledge of its experienced employees.

“We hope that more construction companies will start to think like that and not get caught by surprise when their best people leave and they realise they didn’t write anything down,” he says.

Carpenter is not only talking about companies involved in commercial construction. He is referring to those delivering infrastructure projects too. Recipes for core elements of infrastructure projects can also be created on the platform. 

He says: “Let’s say you are building a freeway and you have got a highway overpass with big supports. You can design a recipe for one of those supports and when you’re creating your plan in Alice, you can just keep copying and pasting that recipe.”

Even though developing these recipes demands time and effort, their availability is expected to considerably cut the time required for scheduling in the long term.

Contractors use the same or similar methods of constructing core elements in different projects, which means that schedulers will not have to start a new template from scratch every time.

“When we build two towers, the layout and the overall sequence might be drastically different. But the way that we pour and place formwork and build individual columns or walls is pretty consistent from project to project,” says Baze.

Alice users are able to change the parameters of a recipe to match the requirements of a specific job. 

Carpenter comments: “HDCC is able to take all the recipes that it already created and move them from one project to the next .”

HDCC will continue to build its recipe library with every project it is working on.

Baze says that the business is looking into introducing Alice to other divisions including the one for civil works. He believes that the company can reach a point where recipes will already be in the system to deliver any project, with the only exceptions being where there are unique client requirements.

The other major benefit from capturing and storing knowledge in the recipes is that it could be a way to educate people in earlier stages of their career.

These individuals will be using the platform to schedule future projects, so they will have access to the input of their experienced colleagues. This will be significantly important after the most experienced colleagues retire.

Carpenter says that young people becoming “fluent in Alice can benefit from the knowledge that has already been captured for them”.

He adds: “They don’t have to guess what it takes to build a concrete column. Somebody else has done the initial work and they can learn from that. It does not only educate the new workers, it gets them productive faster.”

Baze concludes: “How do you take someone who’s been in the industry for two or even 10 years and have them understand things adequately? The only way that you can do that is if you start to digitise knowledge. I think that’s something that our industry in general is not very good at.” 

Like what you've read? To receive New Civil Engineer's daily and weekly newsletters click here.

Tagged with: AI Artificial Intelligence

Sign in or Register a new account to join the discussion.