Easy Simulation by Factory People to Understand and Control Complex, Order-Driven, High-Variety Production

There are numerous production systems which are far more complex than production lines and single-constraint systems. It is not always easy and economical to reduce such systems into a single or multiple production lines or a single-constraint system so that lean manufacturing and TOC methodologies can be adopted for efficient production control and management.

 For production control and management, order-driven, high-variety manufacturing systems are usually far more complex than production lines and single-constraint systems. The complexity may still be high for some of those systems which are equivalent to a production line that makes a variety of products one after another in response to customer orders. Order-driven, high-variety manufacturing systems are mostly found in job shops. Many job shop shops which are small in size and revenue face a lot of difficulty with production control and management.

Difficult Production Control and Management

Many job shop shops which are small in size and revenue face a lot of difficulty with production control and management. This difficulty (to control and manage order-driven, high-variety production) arises due to several factors including the following:

 1.      Work orders may have different material requirements, quantities, priorities, due dates and process steps.

2.      At a work station, the cycle time and resource requirements may vary with type of work order

3.      At a work station, resource requirements may depend on the type of work order

4.      For any given order, the estimated cycle time may vary with work station

5.      Process steps for any order may or may not be sequential

6.      Even if process steps are sequential, the sequence may vary with work order

7.      A process may simultaneously need more than one resource of finite capacity

8.      Resources may include multi-functional machines and multi-skilled workers

9.      Resources may have different patterns of available times (like shifts)

10.  Setup time for a process step operation of a work order may also depend on the previous work order in the sequence

11.  Material supply may not be prompt and reliable.

No Easy and Economical Simplification of High-Variety Production System

 It is not always easy and economical to reduce an order-driven, high-variety manufacturing system into a single or multiple production lines or a single-constraint system so that lean manufacturing and TOC methodologies can be adopted for efficient production control and management. This may be one of the reasons why many job shops ignore methods in Lean and TOC for controlling and managing their production while depending on experience, knowledge, commonsense, whiteboards, Excel spreadsheets, ERP software, project management software and scheduling software. But, those job shops are not still satisfied with the solutions they adopted.

 The difficulty to control and manage such production arises from the large known variation that is described above. Salespersons face difficulty to quote rational lead times for customer orders and wrong quotation of lead times carry two types of risks, one involving potential losses of orders and another related to worker stress in production.

Salespersons may increase the difficulty by their way of quoting lead times for orders. It is not always easy and economical to reduce an order-driven, high-variety production into a single or multiple production lines or a single-constraint system so that lean manufacturing and TOC methodologies can be adopted for efficient production control and management. This may be one of the reasons why many job shops ignore methods in Lean and TOC for controlling and managing their production while depending on experience, knowledge, commonsense, whiteboards, Excel spreadsheets, ERP software, project management software and scheduling software. But, those job shops are not still satisfied with the solutions they adopted.

 It is a major challenge to understand the dynamic nature of an order-driven, high-variety manufacturing system for the purpose of controlling and managing production. Many job shop people are pessimistic about efficient control and management of this kind of production and therefore, they often resort to firefighting on shop floor. There is no widespread, authentic evidence that commonsense methods are adequate to efficiently control and manage such complex systems. Methods based on rigorous calculations can be quite helpful for this purpose but such methods will be accepted by industry people only when they are easily implemented with the help of cost-effective, user friendly software. 

Simulation-Aided Production Control and Management

 Scientific simulation of a complex production system can help understand the nature of the system and make better decisions in production management. But, a major practical difficulty with it is to collect sufficient data for simulating the information contained in customer orders, the arrival patterns of distinct orders, operation times on shop floor and uncertain interruptions in operations. For most practical people in industry, such production simulation may look academic or fancy without practical merit. I would also consider it in the same manner if it is to be done on shop floor in a very systematic manner as taught in schools.

Simple, Easy and Deterministic Simulation of Production

If uncontrollable natural variation and uncertainty are moderate in magnitude in a high-variety manufacturing unit, even deterministic simulation of production with judicious buffer creation will be helpful for (1) predicting and understanding the progress of work orders, (2) what-if analysis, (3) for proactive capacity planning and (4) for scheduling, controlling and managing production.

The large known variation in high-variety production is expected to dominate the moderate natural variation and uncertainty in the system so that deterministic simulation is good enough from practical point of view. To perform sensible deterministic simulation, production people need neither expertise on simulation nor experience with powerful simulation software.

Software for Deterministic Simulation of Complex Production

Some versatile, powerful software developed for generating detailed, operations-level production schedules (without violating any constraints) can be used for deterministic simulation of high-variety production without animation. However, such powerful software requires a good amount of data which includes quantities, material requirements, due dates, priorities and routing information of work orders and resource available times. Nowadays, this data is available in ERP systems or in some Excel files in many industries.

 Since this type of software is primarily designed and developed for manufacturing systems, the development of a scheduling model in it for a target production system should be much easier than the development of a simulation model for the same system in simulation software. The scheduling model in the software can be easily modified by the users using the knowledge of production. Such software can be used not only for actual scheduling of high-variety production but also for performing deterministic simulation to understand how work orders flow through the system, know when and where bottlenecks form in the system and make better decisions in production control and management. 

The detailed, operations-level production schedule generated by the software is taken as the trace of deterministic simulation but the schedule and simulation trace are used for different reasons. The schedule is to be implemented on shop floor while simulation trace is to be used for studying and analyzing the dynamic nature of production including the flow of work orders, bottleneck formations, resource utilization patterns, etc. What-if analysis of production is done by comparing traces of two or more simulations with different input datasets. Simulation-based what-if analysis greatly helps efficient capacity planning.

A few powerful, cost-effective software like Schedlyzer which are developed for detailed, operations-level production scheduling are useful for the required deterministic production simulation in high-variety manufacturing.

The author, Dr. Prasad Velaga of Optisol has more than two decades of experience in developing powerful, scientific scheduling solutions for complex, order-driven, high-variety production units.