Materialize.X is using machine learning to disrupt the $300B engineered wood industry

What’s the next $300 billion industry to be disrupted by technology? Wood. Specifically, engineered wood.

For background, engineered wood is the technical name for any wood product (like particle board) that is created by bonding wood chips into different shapes using an adhesive. It’s much cheaper than using a solid piece of wood, and can be used to make anything from an IKEA desk to kitchen countertops.

Materialize.X, launching today at TechCrunch Disrupt SF 2017, has two new products that it thinks will revolutionize the $300 billion-a-year engineered wood market.

Part 1: a non-toxic adhesive

A lot of engineered wood is created using an adhesive called urea-formaldehyde, which has recently been labeled by the FDA as a toxic carcinogen. So now consumer demand, combined with government regulation, is leading to a shift to non-toxic alternatives in the engineered wood market.

And that’s where Materialize.X comes in. The startup has created a patented non-toxic adhesive to serve as an alternative to urea-formaldehyde. Materialize.X plans to license to chemical companies, or engineered-wood manufacturers so they can make the adhesive on site, the method for making this adhesive.

Part 2: machine learning to improve manufacturing

But creating this new adhesive is only one half of what Materialize.X does. The startup also provides engineered-wood factories with software that uses machine learning to optimize how their adhesive is used in the production process.

For example, right now there’s a standard formula for creating engineered wood — you take wood chips, add adhesive and press them together until they are bonded into the shape you want. But this standard formula doesn’t always produce the best results. That’s because it doesn’t account for variables that can change from day to day, like type of wood, temperature/humidity in the factory and even when the machines were cleaned last.

So Materialize.X has created software that uses machine learning to take in all those variables and make slight changes to the manufacturing process that can greatly improve the quality of the final product. Examples of these changes are adjusting the amount of adhesive used or increasing the pressure in the bonding process depending on the variables listed above.

These two products provide the startup with two pretty diversified revenue streams. Some engineered-wood manufacturers may choose to use the new adhesive but stick to their tried and true methods of manufacturing. On the other hand, others may use legacy adhesive solutions like urea-formaldehyde, but use Materialize.X’s machine learning algorithms to optimize their manufacturing process.

And the machine learning optimization can be useful in other manufacturing processes unrelated to engineered wood — right now the startup is testing algorithms to improve production in the steel industry.