Shape Search and then Some: Part One

By

In the early days of the Internet, circa 1993, the nascent world wide web was facing a huge problem – how to find what you were looking for. The distributed architecture of the web meant that the information you were looking for could be residing on a computer in Berkeley or on a server in Switzerland, and there was no simple way for you to find it.

Enter the search engine… and the world changed forever.

Fast forward to today. Engineering and manufacturing organizations regularly churn out prodigious amounts of data distributed across dozens of different systems – design, sourcing, manufacturing, procurement, quality control, life cycle management etc. And people cannot find anything.

We have all heard the all too familiar lament – “Here’s the CAD drawing I need, but where is the related documentation for this” or my favorite, “Didn’t we just make the exact same thing last year? Where are the files for that?”

Managing the Complexity of Doing Business

Business innovation and technology advancements have made companies more sophisticated, more global and more collaborative than they ever were. Complex product engineering with even more complex supply chains are supported by information systems that are designed to keep everything running smoothly. Some of the largest volumes of data in engineering and manufacturing companies are related to product parts. From CAD files and drawings to sourcing information and inventory reports, data is distributed throughout the enterprise and the supply chain. Companies that need to make rapid, informed design and manufacturing decisions must determine whether:

1) a part can be reused, or

2) a part needs to be modified, or

3) a new part needs to be created or purchased.

But in order to make these decisions, designers and engineers first must determine whether the part or something similar to it exists somewhere within the company’s design vault. Extracting the potential value hidden within this mountain of product data requires an efficient and cost-effective means for finding existing design assets to facilitate future product development through design reuse. The challenge lies in locating a single part in a legacy data trove that may contain a million files or more.

Enter the search engine…for the engineering enterprise.

The Challenges of Locating Part Information

Many technologies, ranging from product lifecycle management (PLM) to 2D/3D database search systems, allow queries of legacy data based on the shape of a part’s geometry—this functionality is referred to as “shape search”. However, such search systems generally do not communicate between each other which means the user needs to go to each source separately and find the data they need.

Also, today’s manufacturing organizations require more than just geometry search to support decisions that maximize design reuse and streamline downstream processes.

The Limitations of Shape-Only Search

Search applications based solely on shape have significant shortcomings. Shape search packages typically support geometry searches from within a specific CAD, PLM, or software application, and do not tap into an organization’s extended data trove of product information.

Shape-searching by sketching the part in a CAD system can be imprecise and wasteful. Making a sketch accurate enough to identify similar parts often takes almost as much time as designing a new part. This defeats the whole point of saving time by searching for similar shapes.

Part geometries can be extremely complex making for an unmanageable amount of shape information to search. For example, do you want to find a part that has the same hole configuration, the same plate thickness or the same fillet radius? Shape-only search typically processes all of this geometric information mathematically in order to find a similar part. On the flip side, if the shape signature/outline is loosely defined, the searches are usually faster; you are in essence loosening the tolerance of the search. The advantage is that more “similar” parts will be found, and the discovery of a larger selection of existing possibilities can lead to further creativity and innovation.

Searching on shape alone is analogous to performing a Web search using a single word on only part of the Internet. Because of the narrow set of programs searched, the results will be incomplete and may not necessarily meet your need.

Most importantly, shape only search is limited in scope, packaged for use solely by designers and engineers, and does not support the search needs of other personnel and/or departments, such as procurement, quality control and inventory management that require ready access to this vital information to make informed decisions about manufacturing.

Informed Decisions Demand an Integrated Search Experience

What is  really needed is a tool that finds and gathers all existing part-related information, including similarity, metadata, and semantic-linked documents and related information to shape-search capabilities—through an integrated search experience that mirrors the manner in which popular Internet search engines and user-friendly ecommerce applications operate.

Exalead OnePart is the latest addition to the SOLIDWORKS family. The EXALEAD OnePart search solution addresses the limitations of other available parts search technologies by extending search capabilities to include multiple types and sources of information, including 2D, 3D, and non-geometric forms of data such as metadata, related documents and BOMS.

Locating a part by its geometric shape is only one step in the process of gathering the information required to determine whether a part is usable or not. OnePart ushers in a new part search paradigm by combining standard search capabilities based on the 3D shape of a part, with natural language text, 2D drawings, existing part metadata, component metadata, normalized metadata, calculated metadata, and semantically linked documents and information. This experience provides you with the deepest insight into your enterprise-wide legacy data.

In Part 2, we will explore the cost of creating new parts, the value of reusing existing parts, and who can benefit (Hint: Procurement, Operations, Inventory management, Supply chain, CAD Admins, etc).

Use the comments section below to tell me your data search story.

Watch the video below to see EXALEAD OnePart in action:

Shyam Venugopal

Shyam Venugopal

Shyam is a Product Portfolio Manager at SOLIDWORKS. He has a Master's degree in Mechanical Engineering from the University of Texas and an MBA from Boston University. He has worked as an applications engineer for the microelectronics/semiconductor industry and has been with SOLIDWORKS since 2012.
Shyam Venugopal