Tutorials on Application Software for Job Shop Lean

Typically, the Lean Thinking Process proposed by James Womack and Daniel Jones is implemented in a factory as follows:

1. Specify Value from the Customer’s Perspective
2. Identify a Value Stream
3. Make the Value Stream “Flow” by Converting the Factory from Departments to Product-based Cells
4. Schedule Production in the Value Stream based on Customer Pull
5. Strive towards Perfection

But, in the case of jobshops, the “Lean Toolkit” for implementing the Womack-Jones process needs to be radically changed and enhanced. Why? Because manual implementation of the Womack-Jones process in any high-variety low-volume environment is going to be slow and cumbersome, inadequate, incomplete and often incorrect! Job Shop Lean relies heavily on effective use of application software and IT (Information Technology) to support the deployment of the Womack-Jones process in high-variety low-volume manufacturing facilities as follows:

• Step 2 (Identify a Value Stream) can be done using the PFAST (Production Flow Analysis and Simplification Toolkit) to analyze the product mix of a jobshop to identify part families (each part family is treated as a Value Stream).
• Step 3 (Make the Value Stream “Flow”) can be done using PFAST, STORM (a facility layout software), and SIMCAD PRO (a discrete event simulation software). PFAST lacks capabilities for detailed layout design and performance evaluation of a Value Stream, an individual cell, or the entire factory. This is a significant shortcoming when management is interested in performance criteria such as throughput, WIP, constraint utilization, operating costs, queuing and material handling delays, etc. In addition, simulation allows numerous what-if analyses to be done to evaluate and rank the benefits gained from changes in layout, setup times, cycle times, handling times, constraint utilization, operator-machine assignments, equipment uptime, etc. Since PFAST, STORM and SIMCAD PRO have common data requirements, the integration of these software tools is relatively easy.
• Step 4 (Schedule Production) can be done using FCS (Finite Capacity Scheduling) and MES (Manufacturing Execution Systems) that complement the existing ERP (Enterprise Resource Planning) or MRP (Manufacturing Resource Planning) systems. • Step 5 (Strive towards Perfection) can be done using PFAST to determine how to reorganize any existing jobshop into a hybrid architecture layout that allows Mixed Mode Manufacturing using Cells and Hybrid Cells and a residual Process Layout.

The primary objective of these software tutorials is to demonstrate that it is not impossible to surmount the challenges posed by the complexity of a high-variety low-volume environment. It just takes a change of attitude, the willingness and ability to learn the relevant tools, and industry projects that provide unquestionable proof that these tools can be successfully used in the real world!
 

TIME

TOPIC

8:00 – 9:00 a.m.

 • Breakfast

9:00 – 9:45 a.m.

(Shahrukh Irani, Ohio State University) Production Flow Analysis to Achieve Flow in any High-Variety Low-Volume Manufacturing Facility

Overview of this Presentation: Production Flow Analysis (PFA) is a comprehensive method for material flow analysis, part family formation, design of manufacturing cells, and facility layout design that was developed in the early 70’s. In a typical application, PFA would be implemented in four stages – Factory Flow Analysis (FFA), Group Analysis (GA), Line Analysis (LA) and Tooling Analysis (TA). Each stage helps to progressively eliminate delays and cut operational costs in a progressively smaller area of the factory. PFAST (Production Flow Analysis and Simplification Toolkit) is a software tool that has automated the manual methods of Production Flow Analysis. This tutorial will give an overview of PFAST, such as data requirements, input file formats, and categories of outputs in the standard PFAST Analysis Report. A typical session with the software will be demonstrated. Finally, the tutorial will outline how to use the outputs from PFAST to:

• Assess the aggregate material flow network and focus on key zones of high transportation costs,
• Assess the aggregate revenue flow network and focus on key zones of high WIP costs,
• Assess the “visual”ness of an entire facility using a simple measure of Line of Sight,
• Assess the feasibility of Cellular Manufacturing,
• Assess the homogeneity of the product mix based on manufacturing routings,
• Assess the cheapest and easiest changes in the existing facility layout that would yield significant reductions in operational wastes and delays that increase door-to-door order flow times,
• Assess the linearity of material flows in the entire facility for a large and diverse product mix,
• etc.

Attendees must challenge the speaker to suggest ways in which PFAST could be used for applications of special interest to them!

9:45 – 10:30 a.m.

(Shahrukh Irani, Ohio State University) Product Mix Segmentation

Overview of this Presentation: In the typical Job Shop Lean deployment project, the first and foremost challenge is to select a representative sample of products out of the 100’s, often 1000’s, of active products being produced, that are the focus for further (detailed) analysis of a large product mix. Given a large product mix (P), how does one choose this sample of products using multiple (often competing or conflicting) criteria, such as Demand or Volume (Q), Routing Similarity (R), Sales or Profit Margin ($) and Demand Stability over Time (T). This tutorial will teach a method – P-Q-R-$-T Analysis – that combines existing methods, such as P-Q Analysis, P-Q-R Analysis, P-R Analysis and P-T Analysis. This tutorial will teach how to use the outputs from PFAST to answer critical questions such as:

• How many products simultaneously contribute to about 80% of Total Material Handling Costs and about 80% of Total WIP costs?
• Which products could be produced in cells i.e. how many clear-cut part families exist in the product mix?
• How many business segments – (High Volume, Low Value) vs. (Low Volume, Low Value) vs. (High Volume, High Value) vs. (Low Volume, High Value) – are being operated under a single roof?
• Is a particular product a Runner, Repeater or Stranger?
• etc.

Product Mix Segmentation is an essential prerequisite for JobshopLean deployment projects such as (i) the Quick Start Approach to JobshopLean, (ii) Value Network Mapping, (iii) Cell Design and (iv) Facility Layout.

10:30 – 10:45 a.m.

• BREAK

10:45 – 11:30 a.m.

(Shahrukh Irani, Ohio State University) Feasibility Analysis for Cellular Manufacturing

Overview of this Presentation: Given a product mix with hundreds, even thousands, of (supposedly) different parts, it becomes a challenge to design some number, X, of independent and stand-alone cells. A primary constraint is the significant sharing of common machines, and overlap of machine requirements, between the different cells. This tutorial will teach how to use the outputs from PFAST for:

• Formation of part families using Product-Process Matrix Analysis,
• Identification of shared machines and exception operations,
• Reduction of intercell flows,
• Management of intercell flows when the shared machine/s are monument/s,
• Design of Hybrid Cellular Layouts when independent and stand-alone cells are infeasible,
• etc.

11:30 – 12:15 p.m.

(Shahrukh Irani, Ohio State University) Design of a Flexible Manufacturing Cell

Overview of this Presentation: How do we select the shape for the layout of a multi-product multi-machine cell? It is well known that the U-shape of a manufacturing cell facilitates process synchronization, material transfers, WIP control and multi-machine tending by any operator, especially for assembly operations. But, there are 25 other letters in the Roman alphabet! What is the best shape for a cell producing a family of parts with similar but not identical routings? This presentation will describe a computer-aided method to determine, on a case-by-case basis, the best shape for a high-variety low-volume non- assembly cell. The proposed method achieves the following design objectives:

1. Minimize total product travel,
2. Shape curvilinear material flow paths,
3. Create a circulation space within which each operator can attend to multiple machines, and,
4. Determine logical locations for inventory buffers.

12:15 – 1:15 p.m.

• LUNCH

1:15 – 2:00 p.m.

(Hosni Adra, CreateASoft, Inc.) Simulation of a Multi-Product Multi-Machine Manufacturing Cell

Overview of this Presentation: The design of a multi-product multi-machine cell extends beyond the initial steps, such as part family formation, machine grouping and cell layout. Problems such as planning floorspace requirements to include WIP, lot sizing, job sequencing at constraint machines, operator-machine assignments, etc. need to be addressed. Simulation software provides a digital mockup of an operational cell that gives the analyst the ability to conduct a variety of what-if scenarios when one or more operational parameters are changed. For example, for any given shift, given the product mix, demand for each product and delivery schedule, the simulation model could help to identify the constraint workcenter. Or, the simulation model could be used to determine the workload imbalance and WIP when the job sequence and kanban lot sizes are simultaneously changed for several products. Or, if the absenteeism of one or more operators necessitates changing the current operator-machine assignments, then new walk routes may need to be determined for each operator based on the actual layout of the cell. This tutorial will illustrate how to model and solve several such problems that arise during a cell design project using SIMCAD PRO.

2:00 – 2:45 p.m.

(Gregory Quinn, Quinn & Associates, Inc.) Finite Capacity Scheduling of a Multi-Product Multi-Machine Manufacturing Cell

Overview of this Presentation: A multi-product multi-machine cell is essentially a small jobshop! Scheduling methods such as takt time and heijunka and generic kanban simply do not suit this manufacturing system for numerous reasons: Not every order follows the same route, there are due dates to be met, capacity constraints influence setup-dependent sequencing at constraint machines, etc. Therefore, the day-to-day operation of a multi-product multi-machine cell must start with a reliable Gantt Chart that displays a valid schedule for each shift and every day of the week. This is crucial since materials, tools, fixtures, etc. have to be delivered JIT to various machines to ensure on-time start and finish of the scheduled operations. Operators must know when to execute various responsibilities, such as loading/unloading, setup and monitoring of machines, or inspection and packaging of parts. Also, it may be necessary to schedule the deliveries of parts from the cell to external resources in the facility, such as heat treatment or electroplating, that need to process those parts. Or, suppliers may fail to meet their scheduled delivery dates or delivery quantities. Or, customers could call at any time and want their parts expedited or delayed, maybe even cancel their orders. In addition to these challenges, the cell foreman and operators are always under pressure to meet due dates, maximize utilization of capital-intensive assets and minimize operating costs in the cell. This tutorial will demonstrate how the PREACTOR scheduling software makes it possible to operate any complex manual (or automated) non-assembly high-mix medium-to-low volume manufacturing cell on a daily basis.

2:45 – 3:00 p.m.

• BREAK

3:00 – 3:45 p.m.

(Brian Mayer, Virginia Tech, and Hosni Adra, CreateASoft, Inc.) Comparison of Physical vs. Virtual Shop Clusters for Reorganization of a Ship Repair Facility

Overview of this Presentation: This presentation will describe a pilot project to evaluate new organizational structures for a US Navy ship repair facility. These structures are:

• Organizational Structure #1: A Physical Shop Cluster (PSC), also referred to as a Repair Cell (or Focused Factory), where several strongly-interdependent shops are physically co-located into a single multi-function shop. This organizational structure is Lean because it reduces transportation, queuing, WIP and other operational wastes. But, it is not Flexible because it cannot adapt to changes in product mix or volume, being limited to the family of repair jobs it was designed for.
• Organization Structure #2: A Virtual Shop Cluster (VSC) where the set of strongly-interdependent shops included in a Physical Shop Cluster (PSC) are not physically co-located and those shops remain at their current locations in the existing facility. This organizational structure is not Lean because it retains the existing transportation, queuing, WIP and other operational wastes. But, it is Flexible because it can adapt to changes in product mix or volume.

Using a representative sample of CASREPS (Casualty Reports) obtained from the US Navy’s Southeast Regional Maintenance Center (SERMC), PFAST was used to identify several groups of shops that had heavy communication and material flows between them. Next, a particular Shop Cluster that was suggested by the PFAST analysis was selected for detailed evaluation. Since PFAST does not have the ability to evaluate the dynamic performance of any system, SIMCAD PRO was used to simulate these two competing organization structures and compare their performance with the Current State of this ship repair facility.

3:45 – 4:30 p.m.

(Shahrukh Irani, Ohio State University) Layout Design of a Multi-Product Multi-Machine Manufacturing Cell

Overview of this Presentation: How do we select the shape for the layout of a multi-product multi-machine cell? In the case of assembly operations, it is well known that the U-shape of a manufacturing cell facilitates process synchronization, material transfers, WIP control and multi-machine tending by any operator. But, what is the best shape for a cell producing a family of parts with similar but not identical routings? Well, besides the U, there are 25 other letters in the Roman alphabet! This tutorial will teach a computer-aided method to determine the best shape for a high-variety low-volume non-assembly cell that fulfils the following design objectives:

1. Minimize total product travel,
2. Design material flow paths with only curvilinear shapes,
3. Create a working zone for each operator that allows him/her to quickly circulate among multiple machines, and,
4. Determine logical locations for inventory buffers.

4:30 p.m.