AI and Machine Learning – Today’s Most Powerful Construction Tools

By | 30 July, 2019

Contractors and major construction companies are each trying to achieve broadly the same business objectives, like streamlining projects, mitigating risks (financial and human), saving time and costs, and ensuring that planning and project engineering are accurate.

The tools many use can be very different, irrespective of role in the process. It’s a surprising fact that even some globe-straddling multinational construction conglomerates run most of their daily operations on Microsoft Excel, given the well-publicized inherent inaccuracy of that particular platform. Is it also surprising to find a tiny contracting company in a niche area of construction running a cutting-edge SaaS to ensure its operations are run smoothly?

Project management, engineering, and construction in general involve huge numbers of variables that make ensuring efficient levels of productivity and progress extraordinarily tricky. The number of involved parties combines with the technical complexity of most projects, and then the whole is seasoned with a generous pinch of unforeseen circumstances. Engineers and contractors can end up in a blame game, as spiraling costs and creeping timeframes ratchet up the pressure.

The knight in shining armor can be portrayed as technology, of course, and every software and hardware producer would like to have the industry believe that by parachuting in solution X or Y, the muddled picture somehow becomes clear. Industry professionals would probably differ in their opinions. So, what are the realities that face the building management system purveyors?

Accuracy is normal(ized)

One of the issues that faces the construction industry and its associated roles is that in pure computing terms, it’s very difficult to normalize data. For the non-technically minded, normalized data is information presented in precisely the same way, irrespective of what it consists. Computers and technology are not particularly suited to fuzzy thinking; here’s an example.

  • A project planning team supposes that element A of a large project will take 30 days to complete.
  • A contractor’s records show that a similar body of work to element A was completed, a couple of years ago. The entirety took six months to finish. The breakdown of the work that was more-or-less the same as element A exists, but the essential figures only format is text in the body of an email that was sent to settle an invoice at the time.

The problem here is one of normalization: information exists that would have a significant bearing on project planning, but it would need manually digging out, and copying and pasting.

The best building management systems need to be capable of drawing together, and normalizing data, with as little human intervention (which is costly, and prone to error) as possible.

Quantifying coal-face opinions

While many project engineers and managers are well-experienced in practical fields (having risen “through the ranks,” perhaps), pulling together different views as to outcomes and predictions is an ongoing consultative process. Not every opinion can be heard, of course, but ensuring there are suitable platforms on which to push back on predictions and guesstimates is important.

The framework in everyday use by project planners is usually highly complex and specialized, and often not the best way to gather feedback. Many BIMs (building information management systems) offer a mobile GUI for real-time updates and comments, but it’s a common trap to believe that by putting a technological solution in place, somehow behavior alters. Changes to operational practice need to come first, if necessary, and then be mirrored in the choice of technology deployed.

Controlling the documents

One of Excel’s failings (see above) is the tendency to play a kind of virtual tennis with Excel sheets, with the added complication of anyone able to declare which “ball” is in play.

Here the oft-mentioned cloud can be a useful tool, rather than an abstract computing concept. Centralized “sources of truth” that are designed for multi-user collaboration with proper version control and activity logs are appropriate platforms for construction.

The same granular configuration of access, logging, and editing also makes the same information available to third parties: contractors, secondary and tertiary companies, for example. That means on- and off-site collaborations are made much more possible.

Automate to dominate

Unlike the capture and assimilation of “fuzzy” information, one thing computers excel at is repetition. That’s a bonus on several levels in the construction industry, as not only does repetitive activity undertaken by humans let errors creep in (as minds wander), but silicon can take on those dull tasks and complete them much quicker.

Right across the product lifecycle, automating certain activities can cut costs, save time on site, and help ensure the accuracy of work in the field. From project planning to helping manage HR issues like transient workforces, the latest and best building management and information systems take on, in a highly controlled manner, some of the repetitive routines that have been time and resource blackholes in the past.

Here at Tech Wire Asia, we’ve put together a shortlist of just two suppliers of software and associated systems that are pushing the envelope of what’s possible in this highly complex and competitive industry. With unique takes built on extensive experience in this vertical, we think their offerings may be just what you need to switch away from your older management and project engineering platform.

INEIGHT

Regardless of size or longevity, most construction companies possess a wealth of historical project data that can be used to guide plans for current or future projects. However, many don’t utilize that information beyond individual memory to inform their plans for greater accuracy and predictability, either because they don’t know how to leverage it, or don’t have the resources needed to execute it.

Enter InEight Basis, a pioneering AI-guided software solution that captures this organizational knowledge or memory and makes it available to project planners for use in creating future plans. Through this tool, historical knowledge can be captured and, via artificial intelligence, generate suggestions and scenarios that are most likely to occur in a given plan because the tool knows what happened before.

Project teams can then vet those suggestions and override or revise them as needed based on human experience. The resulting discussion – and sometimes conflict – arising from this collection of team input feeds back into the plan for greater accuracy than could ever be achieved through traditional AI alone.

This weaving in of a human element – where team members supplement generated suggestions with experiential knowledge and input – represents the next generation of AI, wherein artificial intelligence has transcended traditional computer-based guidance to become augmented intelligence.

And through machine learning, with each plan executed, the tool grows “smarter,” making future suggestions even more accurate and highly attuned to team preferences.

This powerful combination of augmented intelligence coupled with machine learning means today’s construction teams have at their disposal one of the most powerful tools ever created to guide plan accuracy and optimize project outcomes.

We’ll be featuring InEight in a series of articles over the coming months – watch this space for links as they’re published. Or, click here to sign up for a demo of InEight Basis online.

PROCORE

Procore has a range of products that cover off most aspects of the operations of even a very large organization, including offerings for project management, construction financials, quality & safety, and field productivity. Its solutions are cloud-based, and at their heart is the emphasis on effective communications – between people, software, devices and companies.

By providing a single source of information and breaking down the barriers between information silos, all pertinent data is available to all, albeit in a highly controlled, granularly privileged basis.

The key to the interconnectivity of ProCore’s solutions is its advanced APIs and numerous interconnections readily available to deploy from its “App Store”. That means just about any common (or not so common) piece of technology in use in construction today can be integrated into one data “pool”. The App Store covers esoteric drone imagery capture applications through to financial packages, databases and project management platforms, CAD and office software.

Mobile apps are usable as data input devices (text, images, video) as well as disseminators of information – timesheets, project plans, work orders, even delivery notes.

Procore’s applications provide the “one ring to rule them all” approach to data centralization, and like many modern platforms, is entirely cloud based.

*Some of the companies featured on this editorial are commercial partners of Tech Wire Asia



READ MORE