Manufacturers’ all important bottom line. Return on investment (ROI)

All the internet is a blaze with talk of “the internet of things” (IoT) and “big data” for manufacturing. If you are not implementing things like OEE, MES or SCADA equipment you are falling behind the curve, but you have to stop and ask, what is the pay off? And more importantly, what will it cost?

Let me start out by clarifying IoT data has been available in the manufacturing domain for many years but is rarely analyzed to the degree now being touted. I have not personally seen an ROI to justify the cost to log and store every photo-eye flicker in hopes of a predictive maintenance program (motor and VFD information sure). I have on the other hand seen numerous studies on root cause analysis via OEE, MES or SCADA systems that have a 90 day ROI or less!
These are the type of projects that need to be taken from the abstract and put into hard facts but this has its challenges. Industrial Control System integrators are rarely eager to throw ballpark numbers at these topics because of the ambiguity and risk, often the conversation between a manufacturer and the integrator starts out like buying a car.

Manufacturer: “How much does this type of system cost?”

Industrial Systems supplier: “How much do you have to spend?”

From a supplier perspective this is a legitimate question but framed rather poorly, this can lead to undue frustration right out of the gate and I am sure you have witnessed similar transactions. When initially scoping a project for ROM pricing the first question(s) should be goal oriented, “what are you looking to accomplish? Tell me about your goals for this project?”

Once the goals are outlined the scope may be established to determine cost allowing for case studies and ultimately the ROI. The return on investment may be calculated using many different contributing factors, be sure not to overlook any of them when determining this value.

·         Labor savings: rework, operations, defects, time study savings

·         Material savings: scrap, defects, expired product

·         Operating costs: Utilization, energy consumption, labor, storage, downtime costs

·         Maintenance costs: preventive maintenance, parts storage and management

If you find yourself in a chicken or the egg situation, meaning without already having MES (or OEE, SCADA, etc.) you cannot break the data down for ROI purposes. In this situation you may be able to use a low ball defect estimate to get the number, here’s how it works:

• You have 22% defects and everyone has a pretty good idea that 14% is due to failure “A” but the fix must be justified.
• 22% defects cost the company $846,000 annually.
• Estimate is 14% due to failure “A” = $538,363
• MES cost = $182,000, so if you are able to reduce failure “A” by just 5% it will pay for the system in 1 year, furthermore if you are able to completely eliminate failure “A” it will pay for itself in less than 6 months.

So in the ROI document use the 5% required to cover the cost because it is plausible and does not overstate the expectations. Also agree to provide an update to your stakeholders on the data 90 days after implementation.  

Manufacturing Directors, Managers, Accountants and Controllers all love a sure thing, it’s like fixing a horse race. Give them undeniable data to support your improvement requests and they will get behind you. Be prepared and stay focused.

Are there blind spots on your factory floor?

So you have a million square feet of factory floor and you spent (or plan too) the capital to collect data from every piece of equipment, low hanging fruit has been picked yet there is still something missing.  How about the factory floor workers? Are you able to address their needs as easily as your automation or do you have vast black holes of lost communication? Areas where people go in not to be heard from for hours only to report they are out of supplies.

If you installed a flexible and scalable OEE, MES, ERP or any other system of the like it is an easy fix. In any case let me outline some very affordable options to shine a light on those blind spots and increase production.

The solution? Ready? …. Install remote kiosks to form a building wide Andon system. This will give you the ability to log requests, response times, track supply shortages, receive feedback AND display it all real-time on the factory floor or supervisory location. Outfitting everyone with tablets is the wave of the future but due to loss and damage I am going to focus on fixed location kiosks.

Before we get into the software piece let’s look at the hardware. The cost of hardware has dropped dramatically as everyone has expected for years. In fact as you will see there are remote terminal/kiosk options for every budget. For the sake of discussion and planning I will use Ignition as the industrial software because it is light enough to run on any of these computers. I also prefer the Linux OS for remote kiosks as it requires less babysitting in the form of updates.  Starting from the low end let’s look at the new Raspberry PI 2. The Raspberry PI 2 might not be ready for the big time but is worth mentioning.

•        A 900MHz quad-core ARM Cortex-A7 CPU (~6x performance)
•      1GB LPDDR2 SDRAM (2x memory)

Base price is $35, then adding a case, wireless and a larger SD card you are at around $75.  Plenty of keyboard mouse options for ~$30 and a $100 20” monitor puts us at $205 per station.  Honestly, chances are you can find some abandon 16” monitors in the IT boneyard for free but we will keep $100 as the monitor price point.

Next we will look at the ComputeStick form Intel, $110 and nicely equipped. Wi-Fi and Bluetooth standard. 

• Quad-core Intel Atom processor, 1.33 GHz
• Pre-installed with Linux and 1GB memory
• 8 GB of onboard storage
• Micro SD card slot for easy access to additional files
• Built-in Wi-Fi & Bluetooth

Intel Compute Stick $110, Monitor $100, BluetoothKeyboard and Mouse $30, will bring us to $240 per station. I should mention Google has a Chromebit due out this summer that you may find attractive as well.

Okay, you get it, lot of options, so moving on to display all this information on the floor. Starting with the Cadillac of Industrial Displays here is the Arista65” available with or without an onboard Industrial PC. Also available in a touch screen. If that is too much monitor for you HopeIndustrial Systems has a very affordable 23” Monitor for $1080.

For the software piece as I mentioned Ignition from Inductive Automation has the flexibility to create and deploy unlimited remote clients with custom menu’s based on location or login. Meaning the receive area could have a button for “Overage” and the shipping area could have a button to indicate “Tape Low”. Log, track and display data for materials, processes or employees all in one place. A totally integrated system with reporting and historical data without breaking the bank.

So to sum up the numbers I will include 5 kiosks and 2 displays.

Pi 2                                          $205                       X 5          = $1025

-Compute Stick                      $240                       X 5          = $1200

Arista 42”                              ~$5950                      X 1          = $5950

-Hope 23”                               $1080                      X 2          = $2160

-Ignition                                  $12512                    X 1          =$12512

Programming                          TBD

-TOTAL using Compute Stick, Arista and Ignition:              $15,872

Depending on your current situation and future needs an affordable building Andon package is well within reach. Good luck and stay focused.

Successful OEE Implementation Quick Start

Implementing a successful Operational Equipment Effectiveness program is just as much a strategic process as it is a technical one.  Developing a good OEE System is not tricky or elusive it just takes cooperation and hard work. Do not underestimate the effect of putting a manufacturing line or processing system under a microscope will have. Clearly define the goals and roles for everyone prior to implementation, the fewer surprises the better.

OEE is a broad topic and I surely cannot cover all the details in a Quick Start but if you are considering an implementation for your manufacturing floor take a look at these 6 Steps, I hope it saves you some time and heartache.

Understand what OEE is.

Really...there are many KPI’s the true form of OEE may not cover. Chances are everyone you speak with regarding OEE will have a slightly different view of what it will deliver and what it needs to be in your organization. Having a clear understanding is crucial to the conversations.

Get buy-in from EVERYONE.

No part of the group or organization should be exempt from taking action to improve OEE scores. You never know what will be uncovered, if one group will not address an issue the system will stall and lose value as other groups will wait for or resent the non-participant. I had an install which uncovered a work around being performed on the factory floor which could only be addressed by the software group. As you can imagine the software group’s first question was “what is OEE?” Needless to say they where not thrilled about having to address an issue brought to the forefront by OEE. 

Get agreement on a defined list of goals.

Okay, everyone is on-board and you have received feedback, now outline the deliverables. Provide sample reports and screen shots. Once again communicating what OEE will and will not provide in a clearly defined list will prove highly beneficial later on.

Start small, define root cause.

One line or even one machine may be adequate to demonstrate the value. Include the root cause event for every Downtime or Reject. This is a big miss I have seen in implementation. In order to cut costs root cause details are omitted, ultimately ending in a system that does not meet the needs of the manufacturer.

Publish the results!

This is arguably the most important step as it forces action to be taken. Create and publish a report which outlines the OEE scores and details the root cause. Daily is best but weekly is also adequate. For example if Quality is 90.5%, group the 9.5% of rejects into the appropriate categories. If you cannot provide root cause do not publish the report, go back and add the root cause logging required.

Rinse and repeat.

Address the highest offenders on the root cause list until all goals are reached. Then expand the program to the entire factory floor.

Sounds simple right? With a supportive team and a little budget it can be fairly painless.

Good luck and stay focused on the goal..