This is part of the series of blog posts we are writing called “Measuring for Excellence”, in which we’ll be exploring the strategies and best practices some of the most successful companies employ to build a culture of manufacturing excellence in their organization.

Of course, being in the data business, we’ll be paying special attention to the manufacturing Key Performance Indicators (KPIs) they track in order to measure their progress and how manufacturing analytics can track them in real time. In this blog, we thought it would be useful to list out the top KPIs used by manufacturing leaders to create sustainable manufacturing excellence. So, what are key performance indicators for manufacturing plants?:

  1. On-Time Delivery
  2. Work Orders Delivered by the Original Schedule Date ÷ Original Schedule Work Orders Due

    What good is a 100% OEE if we don’t have any customers to sell our product to?

    This metric measures the percentage of orders delivered on-time, most often defined by The metric is often tallied monthly for statistical relevance and the aim should be 100% fulfillment.

    The metric is a favorite of ours ever since we learned it was the number one business goal of one of our best customers, a large contract manufacturer. In the CEO’s office, it was listed number one among many other goals on the whiteboard and was circled for extra emphasis.

    When asked about this KPI, he responded:

    “As long as we are hitting this, we are less concerned about OEE, or even machine downtime,” he told us. “What good is a 100% OEE if we don’t have any customers to sell our product to?”

  3. Schedule Attainment
    Work Orders Delivered by the Original Schedule Date ÷ Original Schedule Work Orders Complete
    This KPI tracks how often the production team meets the target level of production and provides is an important way to set performance benchmarks, fine-tune work order delivery time estimates, and make sure that performance issues aren’t causing costly delays. If a manufacturer only tracks On-Time Delivery for instance, often times, issues within the production process itself might be obscured and the change undocumented.One of the keys to ensuring on-time delivery, if tracked daily, this KPI keeps the production team’s eyes on the prize and could eventually be tweaked to track a level of early delivery performance as well.
  4. Total Cycle Time
    This manufacturing KPI measures the time it takes for a customer order to start and finish the entire production process all the way to shipping and represents the full time required to convert raw materials into finished goods from one end of the line to the other. A Cycle time KPI is the average of all cycle times of all orders in a specific period and is generally calculated using Machine Cycle Time.At the core of any plant performance metric is machine cycle time, defined in detail here. This measure of efficiency sets the bar for how efficient a machine is and allows for real-time reporting on that machines performance (on the minute). Each machine should have an Ideal Cycle Time based on the part being produced.When viewed as set of multiple cycles, it can be measured as Cell Cycle Time.
  5. Throughput
    Units Produced / Time
    This manufacturing KPI is the rate of how many units on average a machine, cell or line is producing over time, i.e. 1200 units/minute. While cycle time is the measure of the time it takes between two points, throughput should be monitored in real-time since when throughput decreases it is usually indicative of an issue on the line.Throughput can be increased by eliminating downtime, calibrating machines to run at an ideal cycle time, reducing the number of touches or steps in cycle to reduce shortstops, changing the raw materials or tooling required to produce the good, improving machine maintenance.
  6. Capacity Utilization
    Actual Output / Potential Output x 100
    If a machine is producing goods at an ideal cycle time, it is said to be running at 100% capacity. When running slower or anytime a machine is idle, this percentage will drop, indicating available capacity and slack in the system. A great KPI to understand the facility’s ability to scale production or institute more agile job scheduling.
  7. Changeover Time
    Changeover Time is the time it takes to unload/load, retool, calibrate, and program a new job. Changeover is most relevant when there’s a switch between one type of part to another before a production run. When taken as an average this KPI can help determine which job types and parts might require some reduction in setup time if possible. By tracking changeover time, manufacturer’s can define total cycle times by part, fine tune their estimates, and recognize the need for more operator training, better planning, proactive prep of required materials.
  8. Yield
    Good Parts Produced / Total Units Produced
    The Yield KPI is a measure of quality and performance and is a the heart of production efficiency and profitability. Measuring First Pass Yield (FPY) will identify which processes require substantive re-work which will affect throughput and influence total cycle times, and provide a target of a 100% yield in which no defective parts were produced at all.
  9. Scrap
    Total Scrap / Total Product Run
    Scrap is the discarded or rejected material from the manufacturing process, so it can be a measure of units or volume. Some organizations track defective items as scrap (waste), while others focus on the leftover raw material from a subtractive manufacturing process, but however your organization defines scrap, tracking this manufacturing KPI should be one of the first steps to lowering your material costs, possibly increasing cycle times, and focusing on producing more quality goods.
  10. Planned maintenance percentage (PMP)
    Planned Maintenance Time / Total Maintenance Time
    By calculating the percentage of scheduled maintenance vs. planned maintenance plus all the emergency maintenance required to address breakdowns. This metric is essential for manufacturers to appropriately allocate resources towards preventative maintenance. One rule of thumb established by advocates of preventative maintenance is 85% PMP, in which an organization is targeting less than 15% maintenance time be dedicated to emergency work orders. Since emergency fixes can cost on average 3-9 times more than planned maintenance due to overtime, rushed parts, service call outs, scrapped production, this metric should be stable for manufacturing seeking uptime and trying to lower operational costs.
  11. Availability
    Uptime / Uptime + Downtime
    At the core of most manufacturing reporting is the availability KPI – the measure of machine uptime/downtime. Downtime is by far and away the biggest loss facing most manufacturers today. No matter what industry you are in downtime costs money. Ideally availability should take into account all downtime, making no distinction between whether it is planned or unplanned. In addition, in order to address the issues causing downtime and reduce it, manufacturers need to start tracking downtime reasons so when viewed on a pareto chart, downtime can be analyzed within the context of the machine affected, by operator and shift, and by any other factor on the plant floor.
  12. Customer Return Rate
    Rejected Goods / Total Number Of Goods Delivered
    As a measure of performance, increasing customer returns might indicate a flaw in the production process or a missing step in quality control. The costs of customer returns can quickly escalate due to the rework required and the effort and cost of reverse logistics.
  13. OEE – Overall Equipment Effectiveness
    Availability * Performance * Quality
    OEE is fine and well– and it has its place in the industry, but we feel strongly that many manufacturers track OEE as a panacea, believing that if they achieve a high enough OEE, they are operating at a level of manufacturing excellence – something we fundamentally dispute. We’ve talked a lot in the past about what’s good and what isn’t regarding the metric (so for our extended take on OEE, see here).