Most Engineering Change Orders Are Preventable

By September 3, 2013Design Reuse, Manufacturing, PLM

Analysis of engineering change orders (ECOs) shows that the vast majority are caused by preventable design and manufacturing errors. These range from simple drawing and data entry mistakes to more complex – yet preventable – errors such as designs that are difficult to manufacture and assemble or to service. In an upcoming webinar, I will discuss two industry studies: an industrial equipment manufacturer where preventable ECOs accounted for 66% of the total, and a heavy equipment manufacturer that estimates that more than 80% of the ECOs were caused by preventable errors. (I published another ECO analysis here.)

These numbers are staggering. However, to put a more pragmatic spin on the discussion, it’s not quite realistic to expect that all of these mistakes can be detected and rectified, as I discuss in another blog post. Yet, the cost of design and manufacturing errors and just managing ECOs are so high, that eliminating even a modest portion of them would be highly valuable.

In the webinar, I will cite two examples of companies that adopted a systematic approach to detecting and preventing potential errors early in the design cycle. An electronic manufacturing services company reported 90% reduction in manufacturing line downtime and 50% reduction in scrap, and overall improvement in product quality and customer satisfaction. An automotive manufacturer reported estimated savings of $4.8M per year from improved throughput and preventing rework, reduction in tooling expenses, and savings in warranty expenses.

  • Joe – I would be interested in your thoughts on how lost design intent scenarios contribute to the error count. What I mean with “lost design intent” is a situation where the original design rationale is lost or not readily accessible such that the effect of change is not fully understood. Even in an optimal design environments where initial designs are front loaded, many change processes take an abbreviated route that seems appropriate for small changes yet unknowingly violate a facet of the original design. This seems to occur mostly in change on change scenarios but can also happen in new design – when responsible engineers are changed midstream due to myriad issues ranging from attrition to resource loading. While previously isolated, I think this sort of causality may be compounding due to recent trends. Such as:

    1. Higher mobility (and hence churn) of engineering employees
    2. Outsourcing and subcontracting where design editing is traversing multiple disparate tools sometimes at the cost of metadata fidelity
    3. The popularity of history free modeling which allows changes to a model regardless of the initial design intent with no viable alternative means to preserve said intent in a transparent manner.

  • Pingback: If You Build It (Using Machine Learning) Will They Come? - Joe Barkai()