Setup Family Sequencing: The Changeover Multiplier Every Job Shop Planner Needs

Changeover time is the tax you pay every time you run a different job on a machine. In a high-mix job shop, that tax compounds relentlessly. A turning center with 8 setups per day at 45 minutes each burns 6 hours—a full shift—just on changeovers. Cut that to 25 minutes per setup and you recover 3.3 hours of productive machine time per day, every day.

Setup family sequencing is the highest-ROI scheduling technique most job shops never implement systematically. The concept is simple: group similar jobs together so the transition between them is fast. The execution requires structure: family definitions, a setup time matrix, and a sequencing strategy that respects due dates while minimizing changeover cost.

After 35+ years helping manufacturers optimize their scheduling at User Solutions, we've seen setup family sequencing deliver 30–50% changeover reductions in shops that apply it rigorously. Here's the complete framework.

For broader context on job shop scheduling optimization, see our ultimate guide to job shop scheduling software.

What a Setup Family Is and Why It Matters

A setup family is a group of jobs that share enough tooling, fixture, material, and dimensional characteristics that transitioning between them requires significantly less changeover effort than transitioning between unrelated jobs.

On a CNC turning center, setup families are typically defined by:

• Diameter range: parts machined from similar bar stock or chuck configurations
• Tolerance class: parts with similar dimensional requirements (tight-tolerance parts often need a different touch-off procedure than commercial-tolerance parts)
• Material family: aluminum, steel, stainless, and exotic alloys all have different tool grades, speeds, and feeds—switching between material families requires tool changes
• Part geometry type: OD turning vs. boring vs. threading vs. facing each have distinct fixturing patterns

A practical example: a job shop with a CNC lathe might define three setup families:

• Family A: aluminum, 0.5–2.5" diameter, commercial tolerance (±0.005")
• Family B: steel or stainless, 1–4" diameter, standard tolerance (±0.003")
• Family C: any material, close tolerance (±0.001" or tighter), requiring CMM verification

Switching within Family A might take 12–18 minutes. Switching from Family A to Family C might take 70–90 minutes. A scheduler who can sequence three Family A parts together before moving to the Family C work saves 60+ minutes of changeover time with zero impact on machine capability.

Building a Setup Time Matrix

The theoretical framework only works if you have real data on your changeover times. The setup matrix captures this data.

A setup matrix is a grid where each row represents the job type you're switching from and each column represents the job type you're switching to. Each cell contains the measured changeover time for that transition.

Here's a simplified 3x3 example for a turning center:

Reading this: if you just finished a Family A job and the next job is also Family A, changeover takes 15 minutes. If you switch from Family A to Family C, it takes 75 minutes. If you sequence your jobs to stay within families, you can avoid the 65–75 minute transitions entirely.

Building the matrix from scratch: For 60–90 days, track every setup at the target machine. Record: job just finished (family or part number), job about to start (family or part number), actual changeover time. At 4–6 setups per day, 90 days yields 360–540 data points—enough to establish reliable averages for each [from, to] pair.

The data often reveals surprises. Transitions that operators assume are "fast" sometimes clock in at 40 minutes because they've normalized a long procedure. Transitions assumed to be expensive sometimes turn out to be quick because a clever operator developed a workaround years ago that never made it into standard procedure.

The Mathematics of Sequence Optimization

Once you have the setup matrix, you can quantify the value of different job sequences. This is where the savings become concrete.

Consider a machine queue with 6 jobs: 2 in Family A, 2 in Family B, 2 in Family C. If you sequence them randomly (ABCABC), you pay the between-family changeover cost 5 times. Using the matrix above, a random sequence might look like:

A→B (35 min) → C (65 min) → A (60 min) → B (30 min) → C (65 min) = 255 minutes total changeover

Now resequence by family (AABBCC):

A→A (15 min) → B (35 min) → B (20 min) → C (65 min) → C (25 min) = 160 minutes total changeover

Same 6 jobs, same machine, 95 minutes less changeover time—nearly 1.6 hours recovered. That's enough capacity for one additional standard job per day. Across a 250-day work year, that's 400 hours of recovered capacity on a single machine. At a machine rate of $80/hour, that's $32,000 in recovered capacity value annually—from sequencing decisions alone.

The math is even more compelling when you consider that the recovered capacity reduces queue depth, which reduces lead time, which improves customer satisfaction and competitive quoting as described in our post on job shop quoting accuracy.

Setup Batching Strategy

Setup batching extends the family sequencing concept by intentionally grouping jobs into a batch that runs as a single campaign on a machine before changing over to another family.

In a job shop context, this means: instead of scheduling jobs strictly by due date and processing them one at a time as they arrive, you hold a family's worth of jobs and run them together as a batch when you're about to make a changeover anyway.

Example: your turning center is set up for Family A. You have 2 Family A jobs due this week and 3 Family A jobs due next week. Rather than switching to Family B this week and back to Family A next week (paying the A→B and B→A changeover costs twice), you advance the 3 next-week Family A jobs into this week's campaign and run all 5 Family A jobs before switching. You pay the A→B transition once instead of twice.

The tradeoff: you're advancing next week's Family A jobs, which consumes capacity today that you might need for other work. The decision rule is:

1. The advanced jobs' customers agree to early delivery (most will, because they get parts sooner)
2. Advancing the jobs doesn't cause other jobs to miss their due dates
3. The changeover savings exceed the carrying cost of the early WIP

In most cases, this calculation strongly favors batching—especially when the between-family changeover is 60+ minutes and the within-family changeover is under 20 minutes.

Offline Setup Preparation: SMED Applied to Scheduling

SMED (Single-Minute Exchange of Die), developed by Shigeo Shingo for lean manufacturing, is primarily a tooling and fixtures technique. But its core principle—move as much of the setup work as possible from internal time (machine stopped) to external time (machine running)—applies directly to scheduling decisions.

In job shop scheduling, SMED's scheduling application means:

Pre-stage tooling and fixtures while the current job is running: The operator or setup technician gathers the tooling, fixtures, programs, and materials for the next job before the current job finishes. This eliminates the time between "last good part" and "machine running on new job" that's consumed by gathering setup items.

Sequence jobs to minimize the internal-time delta: If Family B uses the same chuck as Family A but different tooling, the chuck itself can be pre-staged (external) and only the tooling changes internally. This is a sequencing decision: run Family B jobs directly after Family A jobs rather than after Family C jobs (which require a full chuck change plus tooling change).

Document standard setup procedures by family transition: For each frequently-occurring [from, to] transition in the setup matrix, write a standard procedure that specifies the exact steps, their order, and which steps can be done while the machine is still running. Standardization alone typically reduces changeover time by 15–20% on transitions where operators have developed inconsistent habits.

The combined effect of SMED-informed sequencing and pre-staged setup is commonly a 40–60% reduction in total internal changeover time—beyond what family sequencing alone achieves.

How Scheduling Software Implements Setup Family Sequencing

Manual implementation of setup family sequencing—a scheduler mentally grouping jobs and sequencing them by family—is possible and worthwhile even without software. But it doesn't scale to a shop with 15+ machines and 50+ active jobs.

RMDB implements setup families as a core scheduling constraint:

Family definition: Each job has an assigned setup family based on its routing requirements. The scheduler or process engineer assigns family membership at the work order or job template level.

Setup time matrix: RMDB stores the setup time matrix per machine. When evaluating sequence alternatives for a machine queue, the system uses the matrix to calculate total changeover cost for each sequence option.

Constrained optimization: Given a set of jobs with due dates, RMDB sequences them to minimize total changeover time while ensuring no job misses its due date by more than the user-defined threshold. This is a constrained optimization problem that software handles far faster than manual scheduling—what takes an experienced scheduler 30–45 minutes to reason through for a single machine, the software solves in seconds for the entire shop floor.

What-if analysis: When a rush order arrives, RMDB shows you not just where it fits in the schedule but what changeover cost it introduces. Inserting a Family C job into the middle of a Family A campaign adds two major transitions (A→C and C→A). The what-if report makes this cost explicit—useful data for pricing the rush premium.

Measuring Setup Performance: The Right KPIs

Three metrics tell you whether your setup sequencing is improving:

Average changeover time per machine per week: Track this as an absolute number (minutes) and as a percentage of total available time. A meaningful baseline: in an unoptimized high-mix shop, changeover often consumes 15–25% of available machine time. After family sequencing, the target is 8–12%.

Setup family adherence rate: What percentage of actual setups followed the family sequence plan? If the scheduler sequences jobs by family but the floor reorders jobs for other reasons (expediting, operator preference), adherence breaks down and the savings disappear. Track this to identify where sequencing discipline is being overridden.

Within-family vs. between-family setup ratio: Count how many setups per week are within-family transitions (fast) vs. between-family transitions (slow). As sequencing improves, the within-family ratio should increase. A starting ratio of 40% within-family can realistically move to 65–75% within-family with disciplined sequencing.

EDGEBI can automate all three of these metrics from production data, surfacing them in a weekly operations dashboard without manual calculation.

---

---

Ready to recover 30–50% of your changeover time through smarter sequencing? Contact User Solutions to see how RMDB and EDGEBI implement setup family sequencing and changeover analytics. Trusted by GE, Cummins, BAE Systems, and hundreds of job shops for 35+ years.