Intelligent part number is dead. Long live meaningful attributes?

Intelligent part number is dead. Long live meaningful attributes?


Part numbers are hard. What started as a simple number and references to a part company usually purchase from a supplier, turned out as a complex data management problem with high level of integration complexity.

Part number is a fundamental element of all processes in manufacturing company – started from design and engineering through the manufacturing, maintenance and services. Part numbers are used across departments of manufacturing companies and globally between manufacturing companies and suppliers.

When drawing boards and cork-boards were a foundation of engineering and manufacturing process management, intelligent part numbers used to link piece of information. Was it ideal? I don’t think so. But it worked for many companies. New database driven software came to improve inefficient processes – MRP, ERP, CAD, PDM, PLM. These systems introduced database-driven approach and new problems.

The nature of database is to use a stable identification system. Intelligent part numbers introduced a high level of complexity to all these systems. The ideally database should use internal ID to identify data. Intelligent part number is not a number that databases can use easily and it is a weak link in a chain of data relationships.

At the same time, the reality of most manufacturing companies is multiple part numbers and variety of intelligent solution to map identification attributes between multiple systems and databases.

Jos Voskuil blog The impact of non-intelligent numbers part numbers came to my attention yesterday evening. He speaks about advantages of non-intelligent part numbers and long term ROI of solutions using such part numbers. At the same time time, he acknowledges the need of human-readable properties that are required to work with non-intelligent part numbers. The following passage speaks about that:

Introducing non-intelligent part numbers has it benefits and ROI to stay flexible for the future. However consider it also as a strategic step for the long-term future when information needs to flow in an integrated way through the enterprise with a minimum of human handling.

So when a company decides to move forward towards non-intelligent part numbers, there are still some more actions to take. As the part number becomes irrelevant for human beings, there is the need for more human-readable properties provided as metadata on screens or attributes in a report.

So, king is dead. Long live the king. You can ask what is a difference between intelligent part number and human readable attributes? Jos brings few examples that explains the idea such as multiple classifications, etc.. But at the same time, he acknowledge that a decision to switch into non-intelligent numbers will impact many downstream systems. If you think about non-intelligent number as a database ID, in order to make all systems work, manufacturing company will have to force the same non-intelligent part number to used as an internal database ID in all systems (PDM, PLM, ERP…) or used a unique key to integrate data across multiple systems.

A potential solution can quickly turn into a discussion of standard set of attributes that can be shared by multiple systems. Standard set of attributes will lead us into debates about product data standards, which endless as we know for the last decades. Recently I shared my thoughts about that in my blog – Why standards is not a silver bullet to create a product innovation platforms? which shares some of my thoughts why data integration is hard and standards is not always the best solution.

What is my conclusion? A real problem is data and process integration. The problem is not internal for every manufacturing company, but also comes outside as manufacturing companies are working with suppliers and service partners. Replacing intelligent part number by non-intelligent database ID doesn’t solve a problem, but moves it to the level of systems used to manage information. The introduction of classification schema used by PLM system won’t solve a problem of data management across multiple departments unless everyone will be using the same system. Even imagine company using PLM system to manage everything, the question about suppliers, contract manufacturers, service and maintenance providers is remaining open. It must be a better way. Just my thoughts…

Best, Oleg


Share This Post

  • Sergey Revkov

    Inteligent part numbers (IPN) -это попытка убить двух зайцев , присвоить уникальный идентификатор и создать приемлемое для пользователя краткое описание продукта. Весь пафос проводимых дискусий исходит из колизии между благостью цели и принципиальной противоречивостью подхода. Поэтому уход от inteligent part numbers к non-inteligence parts numbers (NIPN) как индентификационному индексу продукта ,хотя и снимает проблему неприменимости IPN для задачи наличия уникального индекса ,вместе с тем ставит задачу создания краткого описания (а-ля IPN) ,пригодного к использованию широким кругом. Задача облегчается снятием требований по уникальности и неизменчивости данного поля, которое должно присутствовать в представлениях, пользователи которого нуждаются и понимают эту форму и возможностей для расширения не имеется (печатные формы, мобильные интерфейсы и т.д.)
    PS Буду благодарен за перевод поста на английский.

  • beyondplm

    Sergey, thank you for your comment!

    In general, the biggest problem is how to keep both. Another biggest problem is how to manage global identification for parts that in a modern manufacturing environment might have multiple identifications.

    I gave my best shot to translated it. Although I’m not sure I understand it (even in Russian :). If you have future questions, please let me know.


    Intelligent part numbers (IPN) is the attempt to kill two birds with one stone, and assign a unique ID to create acceptable to the user a short description of the product. All discussions is an outcome of conflict between the goodness of purpose and principle contradictory approach. Therefore, avoiding inteligent part numbers for non-inteligence parts numbers (NIPN) as the Record ID index of the product, although it takes IPN Non-Applicability of the problem for the task of having a unique index, however, poses the problem of creating a brief description of (a la IPN), usable a wide range. The task is facilitated by the removal of the requirements for uniqueness and a non-volatile this field, which should be present in the submissions, users who need and understand this form and the opportunities for expansion are not available (printing plates, mobile interfaces, etc.)

  • Sergey Revkov

    We intoduce 6 digits NIPN few years ago. Some inteligence in its creation is that there is no “internal similarity” : i.e. if we have already 123-456 it means that 456-123 or 213-456 are not possible. We place them on all kind of packaging from retail to bulk just near barcodes. In fact -wharehouse people are the main beneficiars of this system . Neveretheless , its are not propagated to retailers -they like something more descriptive for product identification. We suggest something like (USB2100 etc.) The problem arises when we have new product with new different property (not described before) .i.e. we make color variation of product USB2100 >> USB2100 black + USB2100 white. Full of sense adding “black” to our ERP product description doesn’t mean automatic propagation of this change to all chain . So our changed “master data” are changed just in our internal world )). Possible right way is to predict future product’s variation and include de-facto non variable part to description. The probem is that the predictions are unthankfull things.

  • beyondplm

    Thanks for sharing your experience. The integration of identification across engineering-manufacturing-supply chain is essential. The more we move into mass-customization future, the more it will be important.