Plant layout

In one line

There is no plant layout type that is "the best" in general: there are four classic types —fixed position, process, cellular and product— and the one that fits depends on how much volume you make and with how much variety, not on fashion or taste.

What it is

Plant layout is the physical arrangement of machines, workstations, aisles, warehouses and service areas inside a plant. It is not an aesthetic matter: a poorly thought-out layout translates directly into more indirect labor, more travel, more delays from bottlenecks and higher material handling costs. A good layout, on the other hand, is one that manages to make the quantity of product that is needed, at the quality that is required, at the lowest possible cost.

The cited source poses a direct question: is there a layout type that tends to be the best? The answer is no. A layout can be excellent under certain conditions of volume and product variety, and terrible under others. Almost every real plant ends up being a combination of the basic types, applied to different areas according to what each one needs.

What it is for

Choosing the layout type (or combination of types) well determines up front several costs that are very expensive to fix later: how much is spent moving material between operations, how easy it is to train a new operator, how much initial investment is needed in utility lines (air, water, electricity, gas) and how flexible the plant remains to absorb a change in product or volume. Defining this when planning a new plant —or when rearranging one that already works but "grew" in a disorderly way, adding lines as they were needed— is what separates an orderly plant floor from one where everyone walks more than necessary and nobody quite knows why things are where they are.

How it is applied

The four classic types

Fixed-position layout. The product stays still in one place throughout its manufacture, and it is the machines, tools and operators that move to it —the opposite of what happens in any other type of layout—. It appears when the product is too large, heavy or delicate to move: building a ship, assembling an aircraft, installing a turbine or a civil construction project. It fits when demand is low and occasional, the product is highly customized and moving it would cost more (or be flat-out impossible) than moving people and equipment.

Process layout (functional, job shop). Machines are grouped by type of operation, not by product: all the lathes in one section, all the presses in another, the whole paint area in another. Each part travels through the sections that its manufacturing route requires, in the order that applies to it. It gives a plant that looks orderly, makes it easier to train new operators (they are surrounded by coworkers doing similar tasks) and makes it simpler to find competent supervisors, because each one only needs to know one type of equipment. It fits when there are many different products, in low volumes, with frequent special orders. The cost of that flexibility is more expensive material handling (parts travel farther and in different directions) and quite a bit of paperwork to track each order between sections.

Cellular layout. It is a middle ground between process and product. Instead of separating machines by type of operation, the machines needed to make a family of parts with similar process sequences are grouped in the same cell —this prior classification is called group technology—. An operator can run several machines inside the cell, instead of being fixed at just one. The result is less travel and less work-in-process inventory than in a pure process layout, without needing the very high volume that a line per product demands.

According to the cited source, a typical manufacturing cell is a group of six to ten machines run by one or two operators, permanently dedicated to a part or family of parts. Some machines in the cell sit with idle capacity part of the time, but that loss is offset by less inventory, less material handling and less travel time for the part within the process.

Product layout (in-line, assembly line). Machines and stations are arranged in the exact sequence in which the product is made —the assembly line is the clearest example—. It has the lowest unit cost of material handling, because the flow between one operation and the next is minimal by design. In exchange, it requires high, stable volume: if demand drops or the product changes, the entire line ends up underused or has to be rebuilt. It also usually calls for a larger initial investment, because each stretch of the line needs its own utility lines (electricity, air, water) instead of sharing a central installation as in a process layout.

Type	When it fits	Main advantage	Main risk
Fixed position	Large, heavy or highly customized product; low and occasional demand	No need to move a product that cannot be moved	Coordinating many people and equipment moving around a fixed point
Process	Many different products, low volume, frequent special orders	Flexibility and ease of training and supervision	Long travel distances, more paperwork, more work-in-process inventory
Cellular	Families of parts with similar process sequences, medium volume	Less travel and inventory than pure process layout	Requires classifying the families well before setting up the cell
Product	High, stable demand for one or a few products, fixed sequence	Lowest unit cost of material handling	Little flexibility, higher initial investment, repetitive work

Muther's SLP (Systematic Layout Planning) method

Choosing the type is not enough: within each type you have to decide where each machine, each office, each warehouse goes. The most widely used method to organize that decision with data —instead of "by eye"— is Systematic Layout Planning (SLP), developed by Richard Muther. The source summarizes it in six steps:

1. Diagram the relationships. You build a relationship diagram (REL chart) that answers, for each pair of areas or departments, how important it is for them to be close to each other. Closeness codes are used: A (absolutely necessary), E (especially important), I (important), O (ordinary importance), U (unimportant) and X (undesirable —for example, paint next to welding because of fire risk—). Each code can be based on hard data (how many parts or tons move between two areas per day) or on expert judgment when there is no physical flow involved (for example, the closeness between an office and the restroom).
2. Establish the space requirements of each area, in square meters or square feet, based on expected production, the space that already exists or the standards that apply.
3. Build the activity relationship diagram: a visual representation where A relationships are drawn with four short parallel lines, E with three slightly longer lines, and so on, trying to keep the lines from crossing. X relationships are placed as far from each other as possible, joined by a zigzag line.
4. Move on to a space relationship diagram: the same arrangement, but now scaled to the real size of each area, and compacted into a plan.
5. Evaluate alternative layouts. When two or more options seem equally good, you list the factors that matter (flexibility, flow efficiency, safety, room for future expansion), assign them a weight and rate each alternative against each factor. The option with the highest weighted score is the one chosen.
6. Select the final layout and install it.

A simple example from Muther himself, cited by the source, helps show the six steps in miniature: a consulting firm planning to relocate seven office areas (management, engineering, secretarial, reception, files, copying, storage) first builds the relationship diagram (for example, management and secretarial end up as an A relationship, while engineering and reception end up as X, so visitors do not distract the engineers), then calculates the space of each area, builds the block diagram with those relationships and those sizes, and finishes by compressing everything into a final plan. It is a small office, not an industrial plant, but it shows the same procedure used on a full manufacturing floor.

The same logic appears under other names in Meyers' source: the activity relationship diagram (or affinity analysis diagram) is the same REL chart, followed by an intermediate worksheet and a dimensionless block diagram —the first layout sketch, still without scale, which is then adjusted to the real space—. A practical point that this source provides: so as not to overuse the A codes (the most common mistake of someone just starting out), it helps to set a limit —no more than 5% of all possible relationships should end up as A, 10% as E, 15% as I and 25% as O—, leaving the rest as U.

Real example

REENCAVI Compañía Anónima is an Ecuadorian company in the rubber business: tire retreading (the process the company started with), rubber flooring, wheelbarrow tires and granulated rubber. As happens in many plants that grow over time, its production lines were set up as they were needed, with no layout plan behind them — the result was a plant with long, inefficient travel between areas that, in practice, operated with a disorderly process layout: each line added wherever there was space available at the time, not where the material flow called for it.

The study, done with the SLP method, began by diagramming the current travel paths of raw material, work in process and personnel between the areas —receiving, raw material warehouses, the various production lines, shipping, the machine shop— and found that several of those areas (receiving, shipping, the machine shop, the raw material storage, restrooms, the supplies warehouse) were poorly located relative to the real flow of the process, while other sections were in a suitable location and did not need to be touched.

For the rubber granulating area in particular, the analysis of volume and process sequence concluded that a product layout was the right choice: machines arranged in a line, in a "U" arrangement adjusted to the available space, instead of the scattered arrangement they had before. The study calculated that this area needed 145.29 m² —space that the plant already had available, with no new construction—.

The result of redesigning the layout of the analyzed areas: a 30% reduction in unnecessary travel of raw material and stored material, and use of the available space that was at least 40% better than what the previous layout had. The estimated investment to carry out the change (two layout alternatives were evaluated and compared, following step 5 of the SLP) was between 857 and 924 dollars in direct costs, not counting the indirect costs of the adaptation.

The case illustrates the central point of this article well: there was no need to redesign the whole plant. The study identified precisely which areas did need to move (because their location generated long, unnecessary travel) and which were already well located, and for the area with the most volume and a repetitive sequence (granulating) the answer was not "more order within the same scheme", but changing the layout type outright, from a disorderly functional logic to a U-shaped product line.

Template to use

A simple guide to start a plant layout analysis, following the SLP logic:

1. Diagnosis

• List all the areas or departments to be laid out
• For each pair of areas, note whether there is material flow between them and how much (parts, kilos or trips per day)
• Mark the relationships that have nothing to do with physical flow but still matter (safety, noise, need for joint supervision)

2. Choose the layout type (or the combination)

• Is the product large or hard to move? → fixed position
• Is there a lot of product variety at low volume? → process
• Are there families of parts with similar sequences? → cellular
• Is there high, stable demand for a few products? → product

3. Relationship diagram (SLP step 1)

• Assign a closeness code to each pair of areas: A, E, I, O, U or X
• Write down the reason for each A code and each X code (so you can justify it later)

4. Space requirements (SLP step 2)

• Calculate the m² needed per area, with a margin for future expansion

5. Block and space diagrams (SLP steps 3 and 4)

• Draw first without scale (relationships only), then adjust to the sizes and the real space available

6. Evaluate alternatives (SLP step 5)

• If there is more than one viable option, weight the factors (flow, safety, flexibility, expansion) and score each alternative

7. Select and implement (SLP step 6)

So you do not have to build it by hand, there is a flow matrix between areas in Excel with the formulas already done: it adds up the trips in both directions for each pair of areas and sorts the relationships from highest to lowest flow. The values it comes loaded with are examples and must be deleted.

Try it: which layout type fits you?

Three yes/no questions for a first approximation — it does not replace the full SLP analysis, but it helps get the conversation started.

Benefits

• It gives a common vocabulary (the four classic types) to discuss why a plant is laid out the way it is, instead of accepting it as "it has always been like that".
• The SLP method turns a decision that is usually made by intuition or by whatever free space there is at the moment into a process with verifiable steps and data behind it.
• It lets you mix types within the same plant: it is rarely necessary to choose a single type for the whole facility, as the REENCAVI case shows (some areas were kept, another moved to an in-line product layout).
• A good layout reduces up front several costs that are hard to bring down later: material handling, training, supervision, accidents from traffic crossings.

Limitations to keep in mind

• Changing the layout type (especially moving to product layout) usually requires a significant initial investment, including duplicated utility lines that were previously shared.
• SLP depends on the quality of the input data: a relationship diagram built with weak judgment or without real flow data between areas ends up as a layout that looks neat on paper but does not solve the real problem.
• A product layout, even though it lowers the unit cost, is the most rigid of the four: if the product changes or demand drops, it is the one that costs the most to reconvert.
• The selector in this article (and any simple question tree) is a starting point, not a replacement for the relationship diagram or for a real space study — mixed plants or ones with specific constraints (existing buildings, regulations, safety) need the full analysis.

In summary

There is no universally better plant layout type: fixed position, process, cellular and product respond to different combinations of volume and variety, and most real plants end up combining more than one depending on the area. Muther's SLP method —diagram relationships, calculate spaces, build the block diagram, evaluate alternatives and only then choose— is the way to make that decision with data instead of by eye, as the REENCAVI case showed: there was no need to rearrange the whole plant, but rather to identify precisely which areas moved more than necessary and change the layout type only where volume and process sequence justified it.