Avoiding Production Fire Drills

Continuing with our discussion on the Original Equipment Manufacturer-Contract Manufacturer (OEM-CM) relationship, it is important to note that, while many contractor operations are similar, they may have very different procedures and methods for mapping workflow.

The OEM can help by anticipating potential problems and segmenting the workflow to ensure a minimum number of line-down incidents. (See: Talking Points in OEM-Contractor Negotiation.)

Most top-level electronic products share some commonality when it comes to the structuring of bills of materials (BOM). This structure is referred to as a “BOM tree,” and, depending on the selected manufacturing resource planning (MRP) output format, the entire BOM can be represented as “nested” or “indented.”

The lowest level of the assembly is the component level and the unpopulated printed circuit board (PCB). The next level has the stuffed or populated boards, with or without programmed parts. Next level up is the electromechanical assemblies that include mounting hardware, fan brackets, programmed parts, shields, heat sinks, air dams, on-board wire harnesses, and the assembled PCBs.

The top or final assembly usually includes some kind of mechanical housing or enclosure, complete with power supply, wire harnesses, interconnects for points of connections to the outside world, and miscellaneous hardware.

For the purpose of this article, let's assume your model of working with the CM is the consignment version and you are building up to the top level of your product. To facilitate kit and build management, identify and assign a group numbering scheme to the unique assembly structures mentioned above. These groupings are subject to the CM's review — it may already have an established workflow that involves dedicated work centers with bench inventory determined by each work station's function.

The goal is to segregate major stand-alone assemblies to help identify potential build authorizations and to accommodate time-phased inventory awareness. Instead of identifying particular shortages at the piece-part level, potentially stopping the entire product from moving forward in the linear assembly process, it may be sufficient for the CM to say, “Kit group level 1.3 is not complete.” It will then be your call to begin the build knowing that groups 1.1 and 1.2 can proceed as “material complete” while anticipating the timing to fill the shortages for group 1.3.

This will allow time-critical phone conversations without delaying the gathering of detailed knowledge about particular part shortages. The particulars may be determined at a later time and acted upon without incurring unnecessary line-down situations. Below is a suggested breakdown for easy identification of both materials-based issues and also functions or services-based concerns.

  1. Group level 1.1: PCB sub-assembly.
  2. Each board has its own group identifier, 1.1, 1.2, 1.3, etc., including:

    • Printed circuit card (PCB) — 1.1 (Motherboard)
    • PCB components
    • Printed circuit card (PCB) — 1.2 (Daughterboard)
    • PCB components
    • Printed circuit card (PCB) — 1.3 (Power supply controller)
    • PCB components

  3. Group level 2.1: Electromechanical assembly.
  4. Each electromechanical sub-assembly has its own group identifier, 2.1, 2.2, 2.3, etc., including:

    • All sheet metal and machined parts
    • Power supply
    • Fans and blowers
    • All wire harnesses
    • All fastener hardware
    • All cable tie-downs
    • Loctite adhesive other assembly substances

  5. Group level 3.1: Packaging materials and components.
  6. These include:

    • Installation Kit (assembled and part number label attached)
    • Outer shipping carton
    • Set of end-caps (Left, Right)
    • Box labels without serial number label attached

  7. Group level 4.1: In-circuit test (ICT), functional test, and burn-in issues.
  8. This call-out is further identified with the assembly to which it applies, including the following:

    • Parts won't program
    • Ovens non-operational
    • Test equipment malfunctions
    • Software problems
    • Failed ICT
    Examples of service- or function-related call-outs might be:

    • 4.1a/1.1 = Part-programing problem on the motherboard
    • 4.1e/1.3 = Power supply controller failed ICT

By using this method, the entire top-level product is moving via a parallel assembly operation as opposed to a linear, step-by-step arrangement. If there are any shortages detected at the kit audit stage or during assembly due to a high scrap fallout rate, then the unaffected assemblies keep moving forward.

Now your company has a much better system for meeting your delivery schedules as forecasted. There will always be fire drills, but this process of grouping the assembly operations and materials makes the management of those emergencies less harrowing and on-time deliveries more likely.

16 comments on “Avoiding Production Fire Drills

  1. Barbara Jorgensen
    April 27, 2012

    There's really nothing I can add to this, Douglas, but I noticed there weren't any comments yet so I thought I'd start the ball rolling 🙂

    Well said, as always

  2. Himanshugupta
    April 27, 2012

    Douglas, thanks for this post. I did not have an overview of the production assembly but now i have some idea. Your suggestion of grouping the task so that it is easier to identify the shortage and communicate that shortage properly can really work like charm. 

  3. elctrnx_lyf
    April 27, 2012

    This article clearly summarizes how to run a production without getting into to many fire drills. But I didn't clearly understood why the ICT and burnin tests are included in the final package after the complete assembly.

  4. dalexander
    April 27, 2012


    In Circuit Testing, ICT happens post PCB assembly. The functional testing can be at the PCB, Subassembly, or final assembly level. The examples at the end of the article, were for reference only as to how by combining the group ID info, an assembly level and associated problem can be quickly identified and subsequently responded to from a rapid response approach. You can mix and match problem codes, (not all shown) with the various levels of assembly as best makes sense for your product structure and process failure modes. I was able to use this method to keep the line going while part shortages were identified on the non impacted assembly operations underway. CMs have to be meticulous about part audits and process controls and consequently sometime, they will stop an entire operation if anything, anywhere is unexpected. As you may have experienced in a CM line down situation, other customers may slip ahead of you in the queue and that will change the delivery date as originally scheduled. That is what this Grouping ID process is designed to avoid. I apologize for any confusion.

  5. dalexander
    April 27, 2012


    This solution was born out of necessity. When a CM is building a top level product including the shipping materials and packaging, each individual operation requires process-focused attention as there are unique aspects to each operation. By grouping ID names and problems, it avoids miscommunication as to where attention is needed and if there are assigned work centers per operation, a Process or Test Engineer can go to the CM and quickly identify where on the shop floor his/her attention is needed. Thank you for the kind words. It should work like a charm, but it is the adherence to the procedures and good people that make it all work in the end.

  6. dalexander
    April 27, 2012


    I was beginning to wonder if I had missed the mark. Thanks.

  7. _hm
    April 28, 2012

    I agree with you Barbra. This is very good and inforamtive article. However, even after following all these procedures, we do undergo fire drills on occasions. This is due to extreme speed of project and pressure of time to market. Also managing new concept, design, performance achievement and providing all these required documents to CM on-time is very challenging.

  8. Houngbo_Hospice
    April 28, 2012


     “Also managing new concept, design, performance achievement and providing all these required documents to CM on-time is very challenging.”

    I can understand that. That is why this is a team work and everything goes smoothly when the team is well coordinated.

  9. bolaji ojo
    April 28, 2012

    Barb, You always had something to add, even if it was only “well said.” Aside from this, though, from your experience, (and theoretically speaking, of course) can a production fire drill really be avoided? By the way, I'll also be asking Douglas the same question.

  10. bolaji ojo
    April 28, 2012

    Douglas, Is it always possible even after following your detailed prescription for a company to completely avoid a fire drill during or after production? And, if it cannot be avoided, what are the essential steps to taking care of a fire emergency?

  11. dalexander
    April 28, 2012


    I think the best way of handling a fire is exemplified by the city fire department. Since what they do is handle fires, aside from fire prevention through inspection and  code enforcement and approval (Initial Quality Audit), knowing things will  can and will go wrong they:

    1. Expect or Anticipate a fire (be ready)

    2. They have the proper equipment and back-up supplies to combat the various kinds of fire (Every fire has unique characteristics and procedures for mitigating)

    3. They have a cross-trained work force (In case of an unaticipated absence or injury)

    4. They have specific people uniquely skilled for the particular nature of the fire and the complications engendered thereby.

    5. They work as a team

    6. There is a fire foreman calling the shots and keeping the eye on the big picture

    7. They have back up or fall back plans

    8. They can call in extra support from outside their own department

    9. They take pride in their work

    10. They respond rapidly but don't lose thier organization or heads

    11. They stay with it until the fire is out

    12. They learn from each fire and incorporate that learning into subsequent fights

    You take it from there.

  12. _hm
    April 28, 2012

    @Douglas: For this to happen, it needs 11-12 hours a day of work and 10 hrs work on weekend for engineer. After this also you walk very close to fire calls.


  13. Taimoor Zubar
    April 29, 2012

    Also managing new concept, design, performance achievement and providing all these required documents to CM on-time is very challenging.”

    @Hm: I agree that managing component inventory on these guidelines can be time consuming in the short term but I guess it will result in savings in the long run because there will be less glitches in production. Also, it may be time consuming for your engineers if they are not trained to manage the components this way. Once they develop their expertise, they can do it in lesser time.

  14. Taimoor Zubar
    April 29, 2012

    I recently took part in a fire drill at my organization. What they have done is to have a fire incharge in every department who's responsible for ensuring evacuation of all the members in the departments. The fire incharge has a list of all the members under him and he/she cannot evacuate unless all the members of the department have left the building safely. Seemed like a pretty neat setup to me.

  15. Barbara Jorgensen
    April 30, 2012

    Bolaji: I don't think fire drills can be avoided entirely. No matter how well prepared you are — and Douglas seems to anticipate every contingency — there will also be something that goes haywire. I think if you can cover the 80 percent (in the 80 – 20 rule) you are doing pretty well. I've had a couple of situations recently where “I didn't see that coming” — just some bizarre circumstances that must be influenced by tides, the Santa Ana winds, global warming or Murphy's Law. Now, it's funny. But then….

  16. _hm
    April 30, 2012

    @Barbara: I agree with you. 80-20 rule is more appropriate.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.