Lean Tools, Training, and Systems

Lean Flow

When and Where

To learn "What is Lean Flow?"

"Why is it important?"

and "How to make flow happen"

see our other training page for Lean Flow

Why is flow so important?

when

When to design for flow

When to design for flow?

Whenever you can discern a repetitive pattern of any kind.

Flow should be an objective of EVERY production environment

High-volume production enjoys huge cost and efficiency breakthroughs
(as compared to small volume make-to-order)
Perhaps your biggest gains from your lean efforts

will come when you use your Product Family Matrix (vsmProductFamily.xlsx)
to identify patterns that will allow you to organize your products and services into groups that maximize flow

Job shops and service environments
have patterns that will appear when you practice Hansei.

Those patterns can be made to flow.
Projects (one of a kind) can still benefit from most (all?) lean principles and lean methods
Patterns are harder to discern within projects — but is this project really one of a kind?

Might you be able to discern patterns between similar projects?

Pure flow is rare

In real life — you will probably never see a single company with pure flow processes throughout.

There are usually pockets of flow processes, book-ended by batch & queue processes.

A single value stream map might contain sections of:

batch & queue
(e.g. mixing ingredients)
continuous flow
(e.g. to "make goo")
followed by repetitive assembly & packaging.

And the product development and order processing processes for this same product can be designed quite independently of the physical production process.

where

Three lean production environments
that EMBRACE lean flow

Elements of flow can be incorporating into ANY process, but these environments EMBRACE flow

Continuous flow manufacturing (non-discrete products, such as oil)
Repetitive manufacturing (assembly lines)
Lean production (work cells — which can also be applied to service environments)

What they have in common

They embrace the goal

of moving a produced item quickly from activity to activity without interruption for any of the types of muda.

The general pattern depicted in this flow chart

might represent a Lean work cell, a repetitive assembly line, or an oil refinery.

The concepts are the same — 'flow'

What they all reject

All flow processes recoil at the sight of a spaghetti diagram that reveals work flow bouncing wastefully all over the facility

(and getting stuck in piles of either inventory, paperwork, or neglected emails)

in a batch & queue work environment.

Some key differences

	Continuous Flow	Repetitive	Lean
Type of things produced	Food, chemicals, liquids, lumber, paper, textiles, glass, primary metals... Often have lot-controlled products, with coproducts and by-products, and differing potencies or grades	Discrete units, produced in large quantities.	Discrete units, produced in small quantities. Often have product lines with high number of possible configurations. Also ideal for non-inventory service environments.
Primary competitive factors	Commodity price. Quality. Delivery reliability. High yields.	Product design. Quality. Design for manufacturability. Price.	Fast delivery of custom products. Quality. Product design cycle time.
Typical processes	Distillation, heating, reduction, bleaching, grading, spinning, curing...	Assembly, stamping, forging, casting, injecting...	Same as repetitive.
Typical facilities	Many acres, multiple plants, tanks, silos, trains, waste treatment ponds...	Assembly line with balanced takt time per operation, and one worker per workstation	Work cell with balanced takt time per operation. Every person in the cell operates every piece of equipment. Workers "follow each other around".
Scheduling	Focus on utilization of expensive equipment. Scheduled preventative maintenance.	Repetitive Master Schedule for long-term planning. Final Assembly Schedule based on actual customer orders + other demands (such as seasonal build-ups)	Same as repetitive, but very few "other demands". Almost entirely driven by customer orders. Primary objective = flexibility. (which requires "excess capacity")
Process control	Supervisory Data Control & Acquisition (SCADA), distributed control systems (DCS), programmable logic controllers (PLC), statistical process control (SPC), etc. Also Preventative Maintenance (PM), Laboratory Information Systems (LIMS), and regulatory compliance software (e.g. haz mat, FDA...)	Ideally uses Visual Management Systems: e.g. Team Boards, kanbans, andons, poka yoke mistake-proofing... Automated data collection is helpful before and after production, but rarely in WIP. Might use advanced computer process control systems.	Lean uses Visual Management Systems. Theory of Constraints drum-buffer-rope might be helpful. Might use Manufacturing Execution System and automated data collection in production, but the closer to flow, the less needed. Engineering Change Control, Product Data Management, and Workflow software can be important.

Why are work cells the 'holy grail'?

Work cells aren't the only way to achieve process flow.

Assembly lines and continuous process flow lines also work well, too.

However... with production lines we have to address the issues of sales forecasting, batch sizes, finished goods inventories, we have at least removed the gross inefficiencies involved within a classic 'functional department' process environment.

The lean work cell is the 'holy grail' because...

it can accommodate a batch size of one.

And because it can be employed within work environments that have traditionally been organized in batch & queue departments, which have the most opportunity for improvement.

quality

Lean Flow requires (and produces)
higher quality than Batch & Queue

As you can see from this chart — flow processes demand high quality in order to function,
and therefore generate high-quality output, almost as a by-product.

	Batch & Queue	Flow
Effect of defect on work flow	There are usually other batches of work to work on while the defective batch gets reworked or scrapped & replaced	The entire production process screeches to a halt until the problem is resolved (this is called andon or jidoka)
Typical number of defective deliveries before problem is noticed	Usually at least one large batch, and often more than one	One unit (or the smallest possible batch size)
Typical amount of time before problem is noticed	The total time between operations, which can often involve WIP inventory stored for very long periods	Moments
Typical measures of quality	Can be very sloppy. Batch & queue can tolerate poor quality.	Flow cannot tolerate poor quality. Flow processes demand high quality in order to function, and therefore generate high-quality output almost as a by-product.