Featured Product
This Week in Quality Digest Live
Operations Features
Eryn Brown
Their prospects for surviving the pandemic may seem dim, but there are some encouraging signs, experts say
Matt Fieldman
A new Manufacturing Extension Partnership grant will fund a major push in manufacturing innovation
Ryan E. Day
Revolutionizing the supply chain with AI
Manfred Kets de Vries
A best practice is to withdraw from arguments and provide matter-of-fact feedback
Corey Brown
A proactive view on workforce training

More Features

Operations News
ProMation announces additional options for constructing motor-operated valves for industrial flow control
Free education source for global medical device community
Inspect nozzle welds using phased array ultrasound testing techniques including ray-tracing, scanner simulation, coverage maps
March 31, 2021 webinar features carbon and alloy steels
New standard for safe generator use created by the industry’s own PGMA with the assistance of industry experts
Rent with flexibility: ASM Factory Equipment Center
Higher quality contributes to higher efficiency and less downtime
Interfacial launches highly filled, proprietary polymer masterbatches

More News

Jason Furness

Operations

Tales From the Real World: Finding the Weakest Link

Reducing downtime in a cumbersome production system

Published: Monday, June 25, 2018 - 12:02

I would like to share with you a tale from the real world. It’s an extract from the book Michael McLean and I wrote, Manufacturing Money (Amazon Digital Services LLC, 2015). It offers an example of the “Five Focusing Steps” to improvement, especially “Step 1: Identify the constraint,” and “Step 2: Maximizing the constraint’s output.” This was a situation where I looked at a problem for months before I finally figured out where the constraint was.

A large grey iron foundry that I used to run, largest in the southern hemisphere at the time, had a massive green sand casting line. At our peak we made more than 3,000 engine blocks a day, along with crankshafts, flywheels, disc brakes, and other automotive components. The line ran from a basement to two stories high and had a few kilometers of conveyor chain, belts, flat bed conveyors, and four massive 20-ton furnaces. Lifting the rate of production out of this system was essential. We used to run the line for two shifts, ideally six days a week if we could get the people to work, and still had shortages. The line ran at about 135 molds an hour and had been commissioned during the early 1960s. The ground did shake when it ran, which wasn’t often enough or fast enough.

Our maintenance department completed a massive study about downtime for each of the many machines and sections of conveyor that made up this massive integrated system. We had data on downtime occurrences in each area, their duration, their causes, and a plan of fixes. During a nine-day shutdown encompassing an Easter weekend, we spent about $100,000 on replacing worn components, modifying downtime causes, replacing switches, lubricating systems, and generally giving the entire system a major upgrade. We were all confident that this was the game changer we needed to lift our performance and break out of the scrambling pattern the team had been in for years.

We restarted production, and the mold line speed stubbornly remained stuck in the mid-130s. My boss asked me, nicely but pointedly, when were we going to see the returns for all the money we had spent. A week later, no change; another week, still no real change. I was at a loss; we all were. At our downtime reduction meeting where we looked at the data every week, we started to ask what was the piece of equipment that shut the line down most often.

We added some sensors to try and see what was really going on (this was 1960s technology, remember, so data capture was nonexistent until we built it), and what we discovered was that the main conveyor drive motors were being turned off and on every half-second or so by some part of the system. The main mold line never reached full speed because it was being flicked on and off continuously. Because this was happening so often, and the sheer mass of the conveyor system meant that momentum kept the whole thing moving, no one noticed that the line was being turned off and on. The most common piece of equipment sending out a “stop the line” signal was the weight transfer machine.

When you pour molten metal into a sand mold, the air that is already inside the mold becomes superheated as it escapes. To keep the mold from being disturbed by the force of this escaping air, a large weight (700 kg) was placed on top of each mold before the metal was cast. After the metal was poured, this weight was removed and placed onto another mold. This process was enabled by a simple, if rather robust, “pick and place” piece of machinery. This piece of equipment was what was actually stopping the line.

We were all confused about why this machine could be the problem. It never showed up on our manual downtime logs as a significant cause. The maintenance group and the production supervisor went to investigate. The machinery was pneumatically powered (using compressed air), and there was a large 2-in. air line going to the machine. At some point previously a rubber air hose that connected the main line to the machine must have broken. It appeared that no 2-in. hose had been available, so a series of reducing nipples had been fitted, and it was actually a 1-in. air hose that fed the machine, which gave it about one-fourth of the air it should have had. It was starved of air and was running slower than it should have been. No one had noticed. That night we replaced the 1-in. hose and all of the nipples, and fitted a 1.5-m length of the correct 2-in. diameter hose.

The result was instantaneous. The mold line rate jumped to about 145 during the next month, and we peaked at a monthly rate of 170 before I left. With the motor allowed to actually hit full speed, the noise and speed of the system jumped dramatically. All of our downtime reduction efforts became focused on what was stopping the motor from running at high speeds.

During the next few months, we gradually went to a five-day week then a nine-day fortnight because we could reliably run more production with less cost and time. Our weakest link was actually a piece of 1-in. hose.

The $100,000 did give us value because these items needed to be fixed to support the increased speeds. But it wasn’t the weakest link at that time, and I wish we had fixed the hose first.

Discuss

About The Author

Jason Furness’s picture

Jason Furness

Jason Furness, CEO and founder of Manufacturship, is an executive coach who provides lean manufacturing training and lean consulting in a pragmatic, hands-on way that gets clients results in a fast and sustainable manner. Furness oversees the development and delivery of Manufacturship’s curriculum, leads the mentoring of business owners and managers, and sponsors all client projects. During his 20-year career he has led 30 transformation projects for small and medium-sized enterprises. Furness is the co-author of Manufacturing Money: How CEOs Rapidly Lift Profits in Manufacturing (Amazon Digital Services, 2015).

Comments

Listen to what the machine is telling you

I suspect that the sound of the system also changed. It always is a good idea to become familiar with the sound of the system. A good operator will often become aware of a problem in his machine just by a subtle sound shift. In a former position, I had two major issues that were resolved by sound alone. In both instances, it was the "rhythm" of the system that was the clue. Every shop is a symphony and a critical ear can discern that one flat note or tempo error.

Sound

That's true. When you work in a shop long enough you know how it "sounds." If the sound changes, you might not know why it changed, but you know that something is up.

It's kind of like your car. You immediately know if it doesn't sound or feel right. A shop is just a car engine that you work inside of.