Save Hours in Production: Scheduling with Sequence-Dependent Setup Times

In modern manufacturing environments, processing different products back-to-back on the same machine is one of the most critical challenges in production planning. Every product changeover requires the machine to be reconfigured, and when these processes are not managed properly, idle times, capacity losses, and planning errors can arise on the production line.

Sequence-dependent setup time management enables production planning systems to model these transitions in a smarter way. Different setup times can be defined based on the transitions between products, and the system optimizes the production sequence by taking these times into account. As a result, production plans become not only theoretical but also applicable and realistic on the shop floor.

Thanks to this approach, companies can:

• Increase machine utilization rates
• Reduce unnecessary waiting and downtime
• Improve the accuracy of production plans
• Achieve shorter production cycle times
• In particular, within the APS system, it optimizes production transitions by taking sequence-dependent setup times into account and ensures that the production line runs more efficiently.

Application in APS

To apply sequence-dependent setup times in APS, the setup codes and setup times of the products are defined first. Then the setup-time changes that will occur during transitions between products are determined.

Thanks to these definitions, the system automatically calculates setup times based on the production sequence. For example, at a given workstation the setup time may increase when moving from setup code A01 to B01, while the setup time may be shorter when moving from B01 back to A01.

With this approach the production sequence is optimized, unnecessary setup times are reduced, and machine efficiency is increased.

In the system, the standard setup time of code B01 is defined as 3 hours, and the standard setup time of code A01 is defined as 2 hours.

• Scenario 1 (Non-Optimized Sequence: A01 → B01): If the system first ran A01 and then switched to setup code B01, 2 hours would be spent on the standard setup of A01. Because the transition from A01 to B01 is difficult, the system would assign a 4.5-hour setup time for B01. In this case the total setup time would be 6.5 hours.
• Scenario 2 (Optimized Sequence: B01 → A01): The system changes the sequence and processes B01 first, then A01. A 3-hour standard setup time is applied for B01 (See Job 8.1 — Assigned/Total setup time: 3/3 hours). Because the transition from B01 to A01 is advantageous, the system reduces the standard 2-hour A01 setup time to 0.5 hours (See Job 2.1 — Assigned/Total setup time: 0.5/2 hours). In this case the total setup time becomes 3.5 hours (3 + 0.5). 
• The summary screen above shows the total gain obtained in setup times as a result of scheduling. Although the total setup time defined in the system is 73 hours, as a result of the optimized scheduling performed by Dinamo APS the machines were assigned a total of 18 hours of setup time.

This provides a 55-hour setup time saving within the production plan. This result shows that setup times can be significantly reduced when sequence-dependent setup times are defined correctly and the production sequence is optimized.

What Is the Dynamic Setup Transition Matrix?

In production planning, machine setup times are a critical item that directly affects the efficiency of the schedule. In the classical approach, a fixed setup time is defined for each job: on a CNC machine, the operator spends 30 minutes setting up every time a new job starts. However, the reality on the shop floor is much more nuanced. If two consecutive jobs use the same material and the same die, the setup is almost zero; if a different material or a different quality is involved, the time multiplies.

The Dynamic Setup Transition Matrix inside APS was designed to model this reality. During scheduling, the system looks at the properties of the previous job for each new job and calculates the setup time accordingly. The result: jobs of the same type line up back-to-back automatically, and machine preparation time is reduced to a minimum.

Why Is This Feature Needed?

Consider the following scenario: there are 8 jobs on a CNC milling machine. Four of them process CRS steel sheet, and four process stainless steel sheet. A standard tool is used for CRS; when switching to stainless, a different tool, different RPM, and different coolant are required. The operator reconfigures the machine completely at every transition, losing 45 minutes.

With a static approach, 45 minutes of setup is recorded for every job. However, if the CRS jobs come back-to-back, all of them except the first only take 10 minutes.

It is not possible for the planner to do this sequencing by hand; with dozens of orders, dozens of machines, and capacity constraints, this is an optimization problem that must be computed. This is exactly where dinamoAPS+ comes in: you define the rules, and the system builds the optimal sequence by itself.

Classical vs. Dynamic Approach

Both approaches rely on a setup transition matrix running in the background. The difference lies in how that matrix is built.

In the classical approach, the planner enters every transition between defined setup codes one by one. As product variety grows, the matrix expands rapidly and becomes harder to maintain.

In the dynamic approach, the system compares the parametric attributes of jobs (material, mold, color, etc.) and calculates the transition time on its own. The planner defines the rules once; the software handles the rest.

In short, the real difference is who builds the matrix.

How Does the System Work?

The dynamic setup system is built around four different steps. Each step defines one piece; when they all work together, the scheduling magic appears.

Setup Steps

Definition	Function
APSJOBS00	Job list — the setup code and reference texts of each job
APSPROP02	Sequence-dependent transition matrix — code-to-code conversion multipliers
APSPROPMT03	Filter — which machines the rule is applied to
APSPROP03TD	Dynamic code calculation — which fields are read

Flow Chart

When the job list is loaded, APSPROP03TD reads the Reference Text 1 and Reference Text 2 fields in the job records. APSPROPMT03 determines which machine the rule is applied to. APSPROP02 then applies the multiplier for the transition between the previous job and the next job. As a result, the actual setup time is calculated for each job and reflected in the schedule.

Step-by-Step Setup

Step 1 — Setup Code and Reference Texts in the Job List

Three fundamental fields are filled in each job record in APSJOBS00:

• T_SETUP_CODE (Setup Code): The combined form of the setup code and the raw-material code. Example: A01_HM1
• T_REFTEXT01 (Reference Text 1): Only the setup code. Example: A01
• T_REFTEXT02 (Reference Text 2): Only the raw-material/product code. Example: HM1

• Visual 1 — APSJOBS00 job list screen. In each job record the T_SETUP_CODE (e.g. A01_HM1), T_REFTEXT01 (only the setup code: A01), and T_REFTEXT02 (only the product code: HM1) fields are filled in together. The combined code is used for exact-match checking, while the split codes are used for partial-match scenarios. 12 job rows are visible in the list; each row contains machine group, priority, setup time, and assignment information.

The purpose of this triple structure is as follows: when the system performs the transition calculation, it can also check matches based only on the “setup code” or only on the “product”. The combined code (A01_HM1) indicates an exact match, while the split codes capture partial matches.

Step 3 — APSPROP03: Dynamic Setup Code Block

APSPROP03 acts as a general manager: it specifies how the dynamic calculation will be combined. In typical use, a single “AND” block is defined, and the details beneath it are specified inside APSPROPMT03 and APSPROP03TD.

Visual 2 — APSPROP03 Dynamic Setup Code block. At the top of the screen the document number and the “DYNAMIC SETUP CODE” label are shown. In the middle section there are three tabs: Alternative Machine/Labor Group Matrix Definitions, Sequence-Dependent Setup Transition Matrix, and Dynamic Setup Code and Setup Time Transitions Definition. The rule row is defined in the bottom table; the “Independent Block Structure” is set to AND, and the rule definition is active for all jobs.

Step 5 — APSPROP03TD: Partial Match Reward

This is the most critical definition setting. Here, the system learns how to read the Reference Text fields and how much reward to assign to partial matches.

If it is desired to limit the definitions to a specific workstation, filtering can be applied by selecting the relevant workstation (T_WORKSTATION) from the ‘Rules on the Job List’ tab.

Three-Scenario Rule

Rule	Variable 1	Variable 2	Multiplier	Meaning
Row 1	T_REFTEXT01	T_REFTEXT02	0	Both setup and product are the same → standard time
Row 2	T_REFTEXT01	—	0.5	Only setup is the same → time is halved
Row 3	T_REFTEXT02	—	0.7	Only product is the same → time is multiplied by 0.7

These three rules work hierarchically: the system first looks for an exact match, then for partial matches. If none of them apply, the APSPROP02 transition matrix kicks in and the calculation is performed with larger multipliers.

Practical Example: Machine 20CNC01

Now we put what we described above into a concrete scenario. On a CNC machine there are six different jobs, each with a different combination of setup code and product. Let’s see how APS sequences these jobs with the lowest total setup time.

Initial Job List

Job No	Setup Code	REFTEXT01	REFTEXT02	Base Setup
Job 11.1	A01_HM3	A01	HM3	2 hours
Job 14.2	A01_HM3	A01	HM3	2 hours
Job 8.1	B01_HM3	A01	HM3	3 hours
Job 3.1	B01_HM1	B01	HM1	3 hours

Visual 4 — Practical example: scheduling screen on machine 20CNC01. On the left, the workstations list (Guillotine, CNC, Subcontract CNC, PUNCH, Press Brake machines) is shown. On the right, in the Job 11.1 detail panel, Setup Code A01_HM3, planned start/end dates and assigned/total setup time 2/2 hours are displayed; since this is the first job in the sequence, the full setup time has been applied.

After Optimization

APS automatically arranges jobs with the same setup code and the same product back-to-back. As a result, between consecutive jobs either an exact match (multiplier 0) or a partial match (0.5 or 0.7) kicks in; a full setup is almost never performed.

Sequence	Job	Code	Match With Previous	Applied Time
1	Job 11.1	A01_HM1	First job — full setup	2.00 hours
2	Job 14.2	A01_HM1	Setup code and product the same (2×0)	0 hours
3	Job 8.1	B01_HM3	Setup different (3×0.7)	1.21 hours
4	Job 3.1	B01_HM3	Product different (3×0.5)	1.5 hours

Gain Analysis

If optimization were disabled and the jobs ran in a random sequence, each job would take the full setup time: 2+2+3+3 = 10 hours. With dynamic setup, this time dropped to 4.71 hours. Total gain: 52%, on a single machine, over four jobs. Scaled up to a factory with 40 machines, the daily saving becomes dramatic.

Conclusion

The Dynamic Setup Transition Matrix is one of the most powerful optimization tools in the dinamoAPS+ system. It is configured once, and after that it works in the background during every scheduling run, optimizing the sequence on behalf of the planner.