Skip to main content

There is a fundamental problem in problem solving. Many are extremely complicated, so we build models to solve them. But models are just that, simplified representations, not reality.

An example that happened years ago: a contractor had undertaken to transform a basement into an apartment. The architect insisted on ordering the kitchen cabinets rather than having the contractor do it. The cabinets arrived and did not match, as the architect had ordered them to the dimensions of the plans and was not made to measure. The furious entrepreneur went on to explain that no one in their right mind would do that. Instead you would order the slightly smaller cabinets and then a spacer which could be cut to fit if needed.

The architect made the mistake of oversimplifying, relying on blueprints rather than being in touch with the reality of the physical space. And it was in a rehab in the basement. Large-scale, multi-story construction provides exponentially more opportunities for simplification to explode. Some companies are trying to solve this problem through technology.

Sellers love Avvir, Rebuild, StructureSite, Disperse, Open space, To build, Aware of the site, Doxelleand Scale Robotics offer systems and services that are supposed to help keep construction projects on time and on budget by capturing information and performing analysis.

Capturing information is often done by computer vision, which means applying artificial intelligence techniques such as machine learning to images captured by special cameras and then comparing the current state of affairs at any given time to what you would expect given the detailed construction plans. Software systems can then, in theory, see what has or has not been done, and whether there are any deviations from the plan.

It seems related to automated floor plans and space measurements, only much more complex and demanding. If you can document a project’s current process, review it effectively and efficiently, and catch problems as soon as they start happening, not after they’re so advanced that remediation becomes tedious and expensive, it should be possible to waste less money and time.

The remote nature of scanning and then analyzing the data means that much of the monitoring work can be done in an office rather than having people spending time on the construction site trying to document things by hand.

As with any type of software, it can be difficult to compare products from different vendors because they structure and describe things differently. Additionally, companies often cling to increasingly popular terminology, claiming to offer something they might not exactly be able to. But the potential upside is something any company with big plans could explore.