The Challanges of Scale-able Configurator Systems

posted Mar 5, 2014, 5:41 PM by Rick Smith   [ updated May 5, 2014, 8:02 PM ]
I have often been asked "What is a Scale-able Configurator?" and "Why would you want to do that?"

A "Scale-able Configurator" system employs everything from programming both in visual basic and in the case of Inventor, I-Logic utilizing things like case and if statements, to digital prototyping, extensive mathematics, and require associative parameters between every part and assembly scaling the designs up and down depending on a key parameter or set of parameter to achieve the objective. The configurator part of the solution swaps out parts and assemblies via programming usually from reusable libraries of both COTS (Commercial Off The Shelf) and custom materials based on factors such as operational requirements or environmental conditions.
First, it should be noted that a configurator is not appropriate for just any application. 
While notoriously difficult, involved and complex because this kind of project is all encompassing, associating every piece part, every sub-assembly, every assembly, every drawing and more, it will culminate in the drafting and detailing of each part, assembly, parts list, bill of material etc... the project delivers a unique solution and complete drawing or technical data package from the same set of files time and time again like a factory saving time, money and man power over the long term thus making it a worth while investment.

Yes, library parts and assemblies can be created for a unique solution and used to copy and reuse enabling some time savings where in the meta data provided for a project is entered and re-entered to every part and assembly detailing project specific information, revising unique part numbers specific to that project, and finally editing dimensions to adjust sizes and parts are then swapped out in the assemblies to achieve the completed designs, however each part and assembly may still need to be detailed again creating a new TDP (Technical Data Package) each time or you might even have associated drawing files to the library parts and assemblies but only through many days, even weeks of rework can the TDP be delivered and then it still has to be reviewed and revised for error correction. 

The scale-able configurator however, provides a consistent TDP that may need some massage just to clean up dimensions alignment or move a leader note to a position where it is not over lapping some other notes or something to that effect but the captured design intent and approved drawing package already exists and does not need to be revised and re-checked for each project. Once the package has been approved the system becomes a "factory" turning out consistent, correct and unique TDP's where in as with any mistakes, edits or omissions and unlike the "copied library" approach, the data entered cascades through out the project with complete parts lists, bill of materials and unique part numbering delivered for each and every project time and again.
Employing top down methodology, two dimensional cross sections of the envisioned machine such as an elevator are developed. Known as the "skeleton sketch" a single assembly or part file contains multiple sketches which are created in layers and associated by parameters. 
One sketch defines key boundaries of a key element like an elevator shaft for instance, the next defining the platform on which the elevator will ride etc... 
Depending on the application, a part file can be used, however, an assembly file may provide maximum flexibility in the development of this "Root" file if it becomes necessary to reference or position parts relative to key aspects of the skeleton, however, you have to be careful not to create circular references.

The difficult and time consuming part of the process comes when mistakes, omissions and edits are made to the skeleton sketches as the results cascade through out the entire project which can be... well traumatic, maybe even catastrophic which is the reason why such projects are not to be undertaken lightly.
Project planning is key to the development of any application for which a configurator is to be developed with a clear understanding of the projects scope, schedule and resources outlined in a requirements document to provide a clear picture of the project objectives prior to beginning any work the results of which must be thoroughly reviewed at key mile stones where in the resultant drawing package must be signed off on at each stage of implementation to reduce error and ensure successful implementation.
A Scale-able Configurator is an up front investment in time, money, and resources which should be considered in the long view of a company's overall engineering and manufacturing strategy. 

Simply put the "Scale-able Configurator is the apex of what a cad system such as Inventor, Solid Works, Pro-E etc... is capable of achieving. 
In the hands of a competent professional the results speak for them selves and I relish the challenges of this kind of work. 
Such systems once completed are capable of delivering the completed technical data package for a complex machine system such as an electric motor or an elevator lift within a matter of hours rather than days, weeks or even months saving time, money and reducing man hours over the long term. 
Depending on the implementation process, such systems can typically start to deliver a return on investment with in 3 to 5 months and begin delivering the full return on the investment within one year as apposed to time, money and man hours spent on the design and re-design of the same machine under a different project name and basically re-inventing the wheel.

In operation, a form is used to enter relevant project data such as title, project name, maybe a project number etc.. as well as appropriate dimensional information pertaining to the overall sizes for dimension variables A, B, C... which in turn are used in mathematical and associative equations along with components selections for operational characteristics unique to that project. Once processed the associated drawing package changes according to unique project specifications for "Project A"... B... C etc... providing the a for mentioned and required BOM and parts lists. The system is then reset and ready to do it again.
This enables a single individual to produce the complete TDP for a project within several hours again, reducing design cycle time and saving time, money and reducing resource requirements.
Documentation is critical, these systems become highly integrated and in addition to the requirements document it is extremely important to document  a "users manual" as well as such factors as key relationships, programming, the numbering system and how it is applied, where and how meta data such as project, title, and description are captured and applied.
 
The first question from management is "How long is this going to take?"
The real answer is as long as it takes... the tap dancing answer is... "Well, it will depend on the scope of the project, ensuring the identification and construction of the all encompassing sketches ensuring from the beginning that no omissions are made, defining how the nested sketches should be related and in what order, the associations between them and the difficulties encountered as they are defined, as well as the needs of any libraries that will be built to support the project for COTS parts and "standard" custom parts, programming, and numerous other factors and unknowns including but not limited too implementing lessons learned along the way that WILL result in re-work of the skeleton sketches which, because of the associative nature of this kind of project will damage and may even require reconstruction of associated parts or may even result in the realization that the complexities of what has been done so far can, should or even must be simplified and result in re-construction of the entire project. 
While this is somewhat time consuming and sounds drastic it is actually much easier to do the second time around because you have identified all the variables, outlined the requirements and more such that while the initial work may have taken several months to research and develop, reconstruction may be achieved in a matter of weeks resulting in significant improvements that are well worth the effort.
So as a result, rather than setting an arbitrary completion date, it is better to break the project up into phases with mile stones, and phased implementation so as to provide a return on investment as the project progresses, making weekly progress reports, extensive testing and most importantly, holding formal review meetings of the drawing packages and sign off on the completion of the phases prior to implementation so that parts don't come back wrong again and again."
Management starts to get a picture of the effort after that but nothing you say can really give them something they can point to and say we will be done by... Which is understandably uncomfortable and why the requirements document is important.

I have now built five configurators for various applications through out my career including an electric motor, a battery pack, a window and door system, one that defined the combinations and permutations of the various chemicals used in a product and an elevator system.  
The electric motor changed size and parts based on amperage, the battery pack based on the amperage, length and diameter of the battery selected, the windows and doors based on ballistic characteristics required and rough in dimensions.
The chemistry configurator was much more difficult, while the results of this did deliver a drawing package, the objective of the project was to deliver a bill of materials for purchasing.
Including containers and packaging the system defined 98 permutations and 457,163,239,672,252 possible combinations resulting in a discrete 983 line Bill of Material for the project. The system was achieved via the use of table driven I-parts and I-assemblies, programming and component swaps. A sphere represented the chemical compounds in the model but the part number, title and description provided the BOM for purchasing. The system was so complex I had to write a report on it.
The most recent application was for an elevator where in it is the same machine being built over and over with a change in sizes and parts used based on operational conditions such as seismic activity in the region.   

Autodesk, Desult Systems, PTC and others all know that these kinds of results can be achieved using there software and rightfully use that as a selling point talking about how there system can be used to create a "scale-able Configurator" for that project (what ever that maybe) quickly and easily making it sound like doing so is not only simple but the results are like magic! 
However, once the sales pitch is over, I have found that customers of these products while excited by the prospects of what these kind of systems can achieve are largely left with "Our product can do that" but no real understanding of what it takes to do it thus setting very unrealistic expectations by management of those like myself who are capable of constructing such a system but know the realities of what such a project entails. 
Which is why I chose this subject as my first Featured Applications article.  
I hope you have found it helpful and informative. If you have any questions please feel free to write me at info@deltades.com
Copyright March 2014

Comments