Wednesday, October 25, 2017

Principle 6: Building A True Demand Manufacturing Factory

Manufacturing Capacity is Set by Distribution Path

There are specific differences between a demand manufacturing plant and a traditional assembly line mass manufacturing plant. The primary profit approaches are very different. A demand plant is built to deliver product from a virtual inventory and to assure a profit by replacing product that is presold at retail price, preventing clearance dumping. In mass manufacturing the buyer produces profit by leveraging the cost through volume and manufacturing efficiency.  The basic personnel organization of the demand plant is in integrated independent modules that move product traffic through a series of control and quality check points. The mass manufacturing plant, on the other hand, produces a continuous line of production ending in the final inspection of a finished product.  In demand manufacturing the emphasis is on individually addressable accuracy instead of common factor volume efficiency.

Product Discipline

After over twenty years of experience and many painful learning moments the AM4U team has generated a flexible blueprint for true demand production. For years the apparel industry product design has been limited by the cost and risk of decoration because of the minimums and technology of traditional printing and dying.  New technologies like Direct-to-Garment (DTG) and Active Tunnel Infusion (ATI) have enhanced digital printing and dying to virtually remove minimums and time to market risks, enabling unlimited apparel decoration. This ability to print or dye whatever and wherever the designer wants reduces sew in colors and can create thousands of designs from the same blank silhouette or common print.
Silhouette discipline can help with easier adoption of demand manufacturing
changes by limiting the number of pattern and sewing issues and expanding
the opportunities for print and color designs that can be tested without risk.
One of the first rules in designing a demand facility is to understand the range of items each production line is expected to customize on demand. The ability to customize on the fly does not open the opportunity to re-purpose production from shoes to dresses for instance but the modular station structure does allows for the ability to use common functions to create a custom product path.  The demand product path and its modular station structure allow ultimate flexibility to adapt to changing demand volume and customize a range of generic product production; however, product silhouette/print discipline is a fundamental key to quality assurance. The production paths below illustrate this inherent flexibility.

Production Flexibility

This ability to rapidly adapt to changes in specific product demand or product category addresses the key structural danger inherent in demand manufacturing.  Since every level of production ultimately depends on consumer demand the structure of the factory must embrace the peaks and valleys of production at a capital equipment and personnel level.  During the research and design period of our development of a fully integrated PAM site we anticipated the possibility of personnel flex based on demand cycles, however the flexibility required after we began actual production required significant additional creativity. It took years of adjustment and trial and error to develop the structure, technology and techniques to deal with the new challenges of leaving the production schedule in the hands of the consumer.  The critical requirements needed for multi dimensional flexibility described in the next pages will provide a path through the minefield that transforms high risk forecast driven mass manufacturing into a high profit, virtual inventory, purchase activated system.

Critical Requirement One: Flexible Coloring Technology

The single most important technology requirement is shifting the sampling and production coloration from traditional mass production spot color to modern digital process color.  Mass production requires the man hours and cost of color separations, custom color mix, dip approval, screen making, registration, and production printing and drying, which, cannot match the change-on-the-fly instant capability to change the print and colors on a digital printing unit.  

Rotary Screen Presses require color separation, special ink mixes, individual
ink mixes for each color, screens are then constructed and installed, then
registered.  Fabric is then printed in repeat based on cylinder size.  When
complete the press is broken down cleaned and the process repeated.

The textile industry and it’s research organizations focus on tuning old technology and the time and money they have wasted on mass spot coloration has blocked much of the integration required for change-on-the-fly coloration.  If they had spent one percent of those resources on process color we would be changing colors and prints on the fly just like the color printer on your desk.  Digital process color printing, like the printing units that businesses use every day, operates with four or six fixed colors and creates color by adjusting the density of the dots of each color.  The variations in density allow your eyes to process different colors from the reflections of light from the different concentrations of dots.  This is how everything from magazines to labels are printed today it is even how your television and smart phone send you images. The problem is this is not how 99% of apparel is colored today, that process hasn’t changed in hundreds of years.  Digital process color can remove the pre press costs and press preparation labor as well as post-printing cleanup. Process color can make a significant difference in throughput capacity as well as cutting and sewing time.  For instance, digital process color can reduce the number of pieces and seams and sewing steps because colors could now be printed in rather than cut and sewn-in.  In addition the ability to change-on-the-fly reduced the employee headcount and the production waste to a minimum. Financially, the impact of producing daily inventory matching demand created weekly billing and a steady stream of cash flow from customers.  One method of balancing the workflow was to create a second stream of income by producing a house brand product for direct consumer orders and capturing retail income for a flexible shift of B2C online business.  This second channel requires a merchandising staff and may require extra head count in the fulfillment section.  A secondary e-tail channel should only be attempted if the demand facility is linked with an online marketing service and product graphics sources.
Digital inkjet have great versatility to print on fabric, donor paper for sublimation
and even directly on shoes and other accessories.  They change colors on the fly
and can have a color pallet of millions of colors.

Other technologies like optical recognition cutting, piece and automated garment handling systems are important but change-on-the-fly digital coloration is the primary technology adoption that enables a virtual inventory and the associated increase in profits.  The shift to process color is will be very difficult, there is hundreds of years of “we do it this way” mentality to overcome, in addition retraining separators, screen makers, printers, designers and merchandisers to a new more efficient streamlined system will encounter more gate keepers, rules, resistance and more delays than getting a five year old to eat their veggies.  In addition to these roadblocks the performance and color durability test procedures designed by the current authorities are not designed to evaluate, support and create care labels for process color and there seems to be little interest in change.
This is a detailed list of the prep and labor actions that drive print providers to
charge surcharges or require minimums.  Ultimately causing clearances
and dumping losses and operating funds risks for retail and brands.

Critical Requirement Two: Accurate and Timely POS Data

Since there is no demand manufacturing without demand, it is critical for manufacturing to have direct access to consolidated POS data.  This data needs to include distribution path protocols and transit times as well as status of retail on hand sales inventory.  One of the biggest problems with demand manufacturing becoming a reliable source for retail replenishment is POS systems that do not consolidate and order in real time.  The speed and accuracy of inventory movement data is a basic component of building an effective and efficient virtual inventory.  Knowing the characteristics and trends of a products movement is critical to establishing a safe and appropriate on hand inventory and a consolidated Days-of-Supply (DOS) for production scheduling.  Product movement mapping is critical for manufacturing and distribution scheduling, as well as establishing on-hand DOS inventory safety levels at every product movement stop between raw materials sourcing and final customer fulfillment.

Logistics Track

Working backward from the retail consumer transaction POS data is used to build two integrated tracks of information.  The first and critical short-term track is logistics.  This track controls, distribution and transportation in the traditional mass manufacturing structure, however; in a demand structure with a virtual inventory the purchase information actually drives the real time manufacturing of the product and all the associated generic stock unit (GSU) inventory levels and supplies.  The critical accuracy of this information and its timely processing and consolidation will determine everything from manufacturing error rates to the matching of on hand inventory with actual product movement and retail profits.  Expanding current ERP, PLM and EDI capabilities will determine the scope of change and retail recovery over the next few years.

Forecast Track

The second track for POS information is planning.  This track is less immediate but just as important for long term profitability.  The difference between a planning track and a traditional forecasting track is mostly about extending real transactional information as a planning tool versus using associated trend projections as a planning tool.  Forecast models in DM are built on formulas derived and populated with actual sale information.  These extensions include details like sizes and search information as well as micro merchandising preference projections.  Demand assumptions are based on much smaller purchase populations when you are operating from a virtual inventory in a DR or PAM environment.  Forecasting will not disappear but it will be reduced in importance as an actual ordering and financing tool. This occurs because of two factors; first the ability to test market real product in real sales transactions is available through PAM and the customizability of a virtual inventory.  Products and trends can therefore be pretested in store or online without the cost of minimums and with the ability to tweak designs on the fly.  Second, the time from manufacturing order to retail sale can be reduced to ten days or less without minimums or large order stocking requirements.  
Currently apparel retailers are averaging selling
about 27% of a products on hand inventory before
clearance pricing begins.  The inventory glut discount robs
their transactions of the after COGS profits
needed to operate their stores.

 This significantly reduces the risk of overstock or under stock which trend forecasting was supposed to predict but, long manufacturing lead times and the additional inventory cost of short adjustment runs has made forecasting a tool to only predict trends but not to set inventory levels. This was further exacerbated by the focus of software producers on reducing the time of design through sophisticated pattern and 3D visualization software.  This availability of fast product development seduced brands and retailers into more offerings produced through the old mass production system.  This glut of product choice without the development of virtual inventories has constipated the retail track, which is forced to sell more products at greater discounts.  This increase in sales without corresponding profits has created an unsustainable retail environment resulting in todays cascade of retail failures.  Monthly apparel sales data from shows retail sales volume up year over year in August 2017 while clearance sales of price reduced inventory drive gross profits down below sustainable operating cost levels.  

Critical Requirement Three: Strategic Physical Structure

The third critical level of flexibility is the manufacturing structure.  The assembly path of the product defines most manufacturing layouts.  In demand manufacturing the assembly path is a major factor but the quality control system is an equally important design requirement.  The importance of efficient quality management is simple math.  A production error in mass manufacturing when the minimum is 3,000 is an acceptable 03% error rate, but when the minimum is one, any error is a 100% error rate.  It’s important to remember there are no “production seconds” in Purchase Activated Manufacturing (PAM).  An efficient assembly path and no product quality rejections define manufacturing efficiency equally.
AM4U developed a team quality control system, which combined assembly path “stations” with quality oversight to produce an efficient and accurate production model. The product and the physical surrounding structure influence all manufacturing layouts and training, but the essential difference between mass and unit production must be designed into the production structure of the factory or the outcome will not be sustainable.  Further detail of the “station” production structure, technology and training will be discussed in Principle 10.Total Quality Manufacturing Depends on Team Building”.

Setting the Production Output

To be successful, Purchase Activated Manufacturing (PAM) for the individual consumer and Demand Replenishment (DR) for wholesale inventory replenishment depends on one key feature, delivery on time.  Sustained quality delivery depends on detailed prior planning and timely accurate purchase information.  In this section the focus will be on the elements of planning and startup.
Planning starts with the marketing plan instead of the traditional business plan.  This sequence is crucial because almost all the decisions about cost, timelines and production are driven by demand.   Establishing a consistent demand structure is critical to a stable manufacturing structure.  Dramatic changes in demand cause dramatic changes in workload and efficiency.  Fluctuations in demand will affect the manning table directly and the sewing module specifically.  Mixing the customer base with both PAM and DR orders allows for the continuous employment of quality sewing in both modular and progressive sewing production lines.  Modular (sample sewing level) skills are more expensive and are usually required for PAM production, but since PAM sales are retail direct-to-consumer they produce a much higher profit margin that absorbs the higher labor cost.  Progressive sewing lines are used to produce DR products at higher volumes, but at lower wholesale price levels.  Progressive sewing skills are determined by machine type and are task specific, and increased volume at lower labor costs offsets income per unit that is significantly lower.  Factory scale is set by balancing enough digital output of parts in one day to match 2-4 days of output in sewing and fulfillment.  The line personnel ratio is 5 in digital production (coloring, printing, cutting) to 11 in sewing (8 progressive, 3 modular) and 3 Supervisors (Station Manager Digital, Station Manager Sewing and Fulfillment, Production Mgr.) total 19 employees per production line.  Production of standard woman’s leggings per 8hr shift is DR 125-150 and PAM 45-55. Since the digital stations are capable of producing up to1000 individual custom leggings unit bundles per 8-hour shift the total production of the integrated mini factory is based primarily on the number of sewing lines.  This is a standard DR/PAM configuration there are a number of variations like manual cutting, piece drops, outside sewing contracting that AM4U staff tested with varying benefits and results.

The Starting Point

Because this is demand manufacturing not traditional capital intensive mass manufacturing the starting point is a single manufacturing line.  One of the differences between a demand plant and a traditional plant is the agility of the capital equipment.  With the exception of a few pieces of dedicated equipment (like specialized sewing machines) all the manufacturing machines and software are capable of change-on-the-fly task orientation.  This means that the capital equipment outlay and footprint specifications for a complete Integrated Mini-Factory (IMF) are a fixed cost no matter what the product offering.  For instance an IMF to demand manufacture fashion dresses is the same cost as an IMF that makes bicycle clothing.  This feature makes the starting point for building a DR and PAM capability and the resulting profit from a Virtual Inventory tied more to the demand level of the marketing plan than the traditional cost and projections of a traditional business plan.  The IMF is constructed based on actual demand and the scaled by adding appropriate modules based on actual purchase volume.  Each independent module can be tasked to produce a different product or can be attached to a different line to support a spike in demand or a technical interruption in production.

Building the Key Modules

Construction of the Customer Demand and Digital Production modules is a function of identifying and connecting tasks by priority.  For instance, not all of the functions listed in the Consumer Demand Module are required to complete Level 1, Catalog Level, of the Micro Merchandising Plan detailed in Principle Three.  Based on the overall marketing plan the acquisition of functions is a progression based on growth.  Some functions and acquisitions shown in red are critical to the initial planning.  By looking at the entire list however, planners can avoid costly back tracking as they progress to level 3, the Individualization Level or the highest merchandising profit level.

The Consumer Demand Module

The Consumer/Buyer module is the source of the demand information that drives the entire IMF.  The accurate and seamless functioning of SKU, volume, schedule and location information will determine the success of the entire system.  Managing information and graphics from diverse retail ordering systems is the most critical and difficult transition point in adoption of demand manufacturing.
Some brand and retail gatekeepers will use color matching
to resist changes to their current seasonal order routines. 
Establishing an initial color match can help reduce
friction to change.
One shortcut that has worked in the past is to work with brands to identify which fabric, design and garments they want pick to begin building the catalog.  Establishing this color and image base allows the manufacturing module to deal with the inevitable little disconnects that occur during startups. Watercolor Flamingo below is an example of the extension of a print and design across a number of garment and fabrics. 
This approach allows POS and buying functions to sync with virtual inventory, scheduling, production and fulfillment to establish efficiency and confidence in a Demand Replenishment (DR) system.

The Digital Production Module

The Production Module is the most difficult to build, because the development of the equipment and expertise has occurred without any cross pollination of overall direction.  For instance the inkjet companies are not concerned with dot gain and situation because their profit comes from selling ink.  The digital design companies are not concerned with feeding the RIP which translates their visual RGB color into printable CMYK color a key component of color matching.  Remember the single overall goal is a working Virtual Inventory, if any vendor or “expert” does not understand that… find someone else!

Taking full advantage of the Virtual Inventory and Digital Demand production still takes some graphics skills in the conversion of design files to printable nested garment layouts.  However, once the graded parts outlines are nested the task of filling with different prints and colors and storing the SKU in the virtual Inventory  is not difficult.  Finding a SKU in the inventory for production can be difficult if you have not established a uniform naming convention in advance.  Just for reference the normal physical to digital inventory comparison is 1TB of digital equals a full 100,000 sq. ft. warehouse of finished apparel inventory.  A key shortcut for building an efficient and accurate virtual inventory is to build a least one white version of every new SKU and size to check the Tekpak and fit before filling the digital version with color and print designs.

Key Planning Questions

What are you making?
Manufacturing scale, path and fulfillment speed is a function of product type, customer interface and distribution lead-time.  Over the years AM4U has developed a dynamic model that calculates capital cost (purchase or lease), manning table, break-even, and profitability risk based on daily projected capacities, sell-through and fulfillment/distribution speeds for different apparel products.  Once you determine what your primary product will be and the secondary products that the production technology and marketing synergy can support the model is easy load.  The process flexibility of virtual inventory, process color and digital cutting allows wide, no minimum, product selection in the fast section of production.  The slower task and machine specific sewing and finishing sections can be fed from a hub of coloration and cutting.  Matching the output per shift of the faster coloration and cutting modules with the slower sewing and fulfillment modules sets the production integration levels between modules. The fixed capital, materials and labor cost of the coloration and cutting modules compared with the variable capital and labor costs of the sewing and finishing modules makes the gross profit and delivery calculation for each order easy to compute, quote and schedule. 

Who are you selling to and How are you selling?

Integrating the production planning with both promotional PAM merchandising and seasonal DR retail customer promotions can both level production and sustained growth.  Balancing planning for hot PAM selling periods like Mother’s Day, Christmas or “back to school” when selling direct to the consumer may require additional personnel, first in fulfillment and finishing and then in the sewing module.  When scheduling overlapping PAM and DR a key feature of planning manufacturing volume based scheduling is the seamless transition between the production period required for DR distribution to retail customers (3-5 weeks) and the production lead times (5-10 days) for PAM orders.  This DR lead time allows part time and/or overtime employees to have the highest level experience and cross-training competence when individual and rush orders arrive.


In short, virtual inventory demand manufacturing integrates retail, brand and manufacturer growth and profit goal into one unified plan.  High profit Integrated Mini-Factories (IMF’s) using clean digital coloration and cutting while pulling product from vast virtual inventories is available today.  Brands are already feeling the pressure retailer’s are feeling today from the unsustainable pursuit of profit through volume leverage versus customer valued product customization and clean manufacturing.  Integrated Mini-factories will impact the apparel and textile segment just like micro brewers and vintners are impacting the beer and wine market.  
AM4U has spent the dollars and time to build the bridges, techniques and technology required to integrate visual design, digital merchandising, virtual inventories, demand driven production, and real time testing of the critical concepts of PAM and DR integrated factories.  
E-mail me at “” and we will help you test demand manufacturing for yourself.