Pro-MACS Return on Investment

By replacing traditional troubleshooting with automated troubleshooting we can increase maintenance productivity.

While these men are trying to diagnose the problem the downtime is accumulating.

There are useful reporting tools, which produced easy to read Production, Downtime and Failure Reports. This way the mechanic can get a precise answer without sifting through overwhelming data.

To further the discussion of Automated Maintenance Diagnostics we must first define the anatomy of a machine failure. The time between the occurrence of a failure and its full repair is divided into four periods.

• Period 1: Alarm time is the time until the alarm is detected.
• Period 2: Location time is the time until the machine failure is found.
• Period 3: Conclusion time is the time until a conclusion is drawn as how to repair or correct the failure.
• Period 4: Repair time (or Wrench Time) is the time to repair the failure.

Periods 1, 2, and 3, are the most critical because this is when the biggest damage to products and machines occur. Let's look at the effects on each time period.

Because the system informed the mechanic immediately with voice, text and graphics, that a failure has happened it reduced Alarm Time to almost zero.

When the system showed the mechanic exactly which machine and part had failed in three-dimensional graphics exactly the way the machine looks in his plant it also reduced Location Time to a minimum.

Because the mechanic was given diagnostic advice with a list of causes and possible solutions Conclusion significantly was quicker.

The knowledge of how to repair the failure and part numbers was made available on a failure information screen. The experience of the maintenance staff could also be added. This helped accelerate the Repair Time.

The results are shown in the following graph: The first three periods were reduced an average of 90% and the fourth period was reduced 20% as shown in the right-hand side of the graph.

Below is the analysis of a bread line producing 7200-7800 loaves per minute.

Additionally if we consider a typical bread line has 80-100 hrs of downtime per year and we further calculate 100 hours times $5,000 per hour, times 34% savings, we have saved $170,000 for the year.

This brings us to Predictive Maintenance; many people confuse it with Preventive Maintenance. First let’s look at the difference between Preventive and Predictive Maintenance.

Preventive Maintenance is the routine of scheduling and performing repair tasks on your equipment before it becomes necessary. Predictive Maintenance, on the other hand, is the process where data about the equipment is collected and analyzed in the effort to predict a machine breakdown or failure.

 

The progress of Predictive Maintenance for the food and baking industry has been slow mostly due to the cost per result ratio. In the past it was less expensive to increase Preventive Maintenance tasks and put more emphasis into hands-on inspection.

It is only recently that the costs of monitoring sensors and PLCs have dropped, whereas the cost of maintenance tasks has gone up. However, the real progress is not in just the sensors, it is also in the analysis software.

The complexity, size and costs of the food and baking equipment has risen, so is the risk of longer breakdowns. Larger machines mean larger, more expensive drives and gearboxes. More automated machines mean more complexity in the controls and equipment. The technical expertise of maintenance is not keeping pace with technical innovations from equipment suppliers.

The Machines to be analyzed must be PLC controlled and additional sensors or devices may be needed. Next a computer program using ProMACS is employed which will communicate with the PLCs. Then Predictive Analysis Programming is added to the PLC programs. The machine parameters and normal operating tolerances must be recorded and input.

Predictive analysis includes many parameters.

February 24, 2005
• Meantime Between Failures
• Deterioration of Individual Machine Operations
• Increase in Vibration Tolerances
• Increase in Motor Amperages
• Changes in Sonic Resonance
• Increase in Component Temperatures

Why Predictive Failure Analysis?

• As an addition to Preventive Maintenance programs
• As an extra pair of eyes and ears that work endlessly
• To prevent unplanned shutdowns.
• To reduce the frequency of Preventive Maintenance tasks

What can we expect from Predictive Failure Analysis?

A lot, but it will only come with the partnership of the plant maintenance personnel in commissioning and working with such a system. You should:

• Invest two half hour periods per week to review collected data and input any manually collected data.
• Expect to invest one half hour period per week for the 1st few months to adjust and tune set points and thresholds.

Then you can expect less unplanned shutdowns.

Where would you apply it first?

• Machines that are the most difficult to repair.
• Machines that have expensive parts like $20,000 gearboxes.
• Critical equipment like unloaders and loaders for ovens and proofers
• Machines that would cause a bottle neck to production

The Costs:

• Computer monitoring hardware & software $6,000
• The programming and hardware for each Machine about $1,500 - $2,000

The Benefits:

• The benefit of scheduling a shutdown for repairs, before you’re forced to do so.
• Cost savings per downtime hour is $5-8,000 for a breadline, $12-16,000 for a high volume cookie line, and for a pastry line $16-24,000.

Independent studies by the ISI have shown that predictive maintenance can reduce preventive maintenance tasks by 15% and have an overall effect on downtime of 1-2%. If we look at the additional productivity of a maintenance staff of 30, the calculations would be as follows;