Identifying and accurately calculating spot elevations — high or low areas on an excavation site — on earthworks drawings can be tricky work for an estimator.
“There is so much going on with a set of drawings that to only find the spot elevations actually means training it (software) to disregard all of the irrelevant stuff on a set of plans,” says Jeff Graham, CEO of Construction AI, a young tech company which is planning to launch software soon that will solve the problem.
It is important information for estimators, particularly on extensive excavation sites such as subdivisions, schools, commercial and industrial malls, pipelines and road projects, says Graham who co-founded Construction AI with Max Klein and Jason Heard.
Calling an estimator’s time “unbelievably precious”, Graham says the software allows them to prepare a bid quickly and accurately.
Filtering out irrelevant information is the trick, he says, noting that to accurately calculate spot elevations includes determining the cost of fill removal and the type and quantity of replacement materials.
While optical character recognition software has been used by some companies to scan drawings, it does not focus on spot elevations, says Graham whose company is based in Mosquodoboit Harbour, N.S. and Alberta.
“Estimators need to figure out the areas of new asphalt, concrete, storm, sanitary, water. It’s complicated.”
Construction AI’s software has been in development since the company was founded about three years ago.
“It’s been a brutally hard problem to solve,” he says, noting that his is not the first company to take a stab at developing the software.
Graham says the software “frees up” estimators to do other tasks. Securing the closest dump sites to keep trucking costs down and managing subcontractor relationships are examples of those tasks.
“I think those are ultimately the things that can win or lose a job.”
Graham says the software will be launched once a “channel marketing” deal has been arranged. That will be “a revenue-sharing deal” with an established earthworks software player which has a global-reaching customer base.
The goal is to form a partnership with “with a construction company or two in Canada.”
Graham says the software is not suitable to every project.
“Our use case is where the engineer does a site model in CAD and then renders it in a PDF drawing and posts it on a tender site” for bidding earthworks contractors. “Each one of those contractors turns that PDF back into their own 3D model or they calculate it by hand using cross sections.”
He says Construction AI can process the PDF quickly to get the information to contractors to accurately bid the job.
“The problem we are solving is in the tendering process where the engineer doesn’t share the CAD file, which in my experience is most of the time.”
The idea of developing the software came to Graham while he was working for his father’s excavating business. Describing the estimating work as “drudgery,” he says it involved entering spot elevations, identifying buildings and other data entry tasks from drawings to exact the quantities for a bid submission.
Construction AI was a recent winner of the Innovacorp Sprint Competition for early stage software-based technology companies in Nova Scotia. The company received $25,000 and guidance from seasoned business professionals to help get the product to market.
Graham says in the near future the software will do more than accurately estimate contours (spot elevations). “We’re very confident we will be able to do roadworks, sidewalks…”
Over the longer term, Graham says the goal is to provide software to do repetitive tasks for other construction trades as well.