Perfection Isn’t Always a Good Idea

In the 1980’s, Thomasville Furniture was the market leader in high-end furniture. They were extremely focused on process improvement in all areas of the business. I worked on a staff of industrial engineers. Each of us supported one or more of the many manufacturing facilities. We were tasked with finding ways to improve safety, quality, productivity, and customer service.

One of my colleagues worked at the plant that made tables and chairs. They took great pride in the finish of their tabletops. To achieve the desired look, many workers used powered sanding and rubbing devices. It was an extremely labor-intensive process. We had many muscle strains as a result from all the hand sanding and rubbing to get the desired look.

We heard about companies using robots to reduce labor intensity and safety risk. It seemed like an opportunity for us. If we could figure out how to get robots to sand and rub each tabletop, we could realize significant labor cost reduction and eliminate one of our top safety risks.

My colleague, Frank, connected with a robot supplier who brought in some of his robots for a 30-day “test drive.” The first few weeks were spent teaching the robots how to find the surface and perimeter of the table, apply the proper pressure, and cover the entire surface of the table consistently.

Eventually, they were able to achieve their objectives. The robot was able to sand and achieve a rubbed finish on the tabletops, without human intervention. The tabletops looked great. We were ready to roll out robots to take over the sanding and rubbing process.

Before fully committing to the robots, they decided to show the results to a focus group, to get their reactions to the results. The feedback was surprising. Initially, the group loved the finish and look of the tables. In fact, they said the finish was better than anything they had ever seen. But, after more review and conversation, there was something that bothered them. A series of questions drew out the issue: the tops were too perfect. There was no variation across the surface. We had lost the “handmade” look and feel of the furniture. Instead, the lack of variation made the tables look mechanically made, plastic, or unreal.

We couldn’t give up the thing that made Thomasville Furniture a market leader. Frank and his team went back to the robot vendor to see if he could program variability into the process. After many attempts, they gave up. The technology at the time wouldn’t allow the robots to randomize the pressures applied across the tabletops.

The team removed the robots and returned to the time-tested approach to hand sand and finish tables. Over time, robot technology improved, but it was never given another chance at Thomasville Furniture. Instead, we worked on ways to simplify and reduce the safety risk of hand finishing tables.