Talk:ICAD (software)

Historical significance
As the aura of numeric modes fades, and the mania abates, this little example will be useful. So, I put a little note in the Talk page of the constraint satisfaction page. We need to balance symbolic/numeric/proof theoretic/etc. modes with the trained intuition brought by the engineer. jmswtlk (talk) 13:13, 13 November 2021 (UTC)


 * Turns out that there has been lots of KBE work done over the past two decades. Some of this was done via embedded systems from 3rd parties. But, here is an example of ICAD use that needs to be used here (KBE Design Methodology, Collegiate Level). jmswtlk (talk) 02:32, 1 January 2024 (UTC)

Status
This is a first attempt. There are plenty of holes to fill.

Anyone with information, please contribute. — Preceding unsigned comment added by 67.66.111.249 (talk) 17:27, 3 November 2005


 * There have been many changes, but we need to mention projects. For me, the Boeing 777 has to be represented. It was a case where KBE was both new and necessary. And, it produced benefits such that subsequent programs want more KBE. Of course, 777 used ICAD (at the time, SUN workstations tied to RSs and mainframes). ... I was still using ICAD in 2005 when I retired on some early modeling for the 787 (perhaps, the culmination at Boeing). jmswtlk (talk) 01:11, 21 August 2015 (UTC)

Lisp Machine, et al
We need to tie in with the history of software as influenced by Lisp and Lisp Machine work. jmswtlk —Preceding undated comment added 15:59, 27 December 2005.

KTI
The 'ktiworld' domain name expired on 01/22/2006. It is 'pending renewal or deletion'.

This article will need to be updated accordingly (reference pointers, etc.) jmswtlk 20:54, 2 February 2006 (UTC)


 * Everywhere there is 'www.ktiworld.com, change to 'ds-kti'. jmswtlk 17:24, 8 March 2006 (UTC)

Greenspun
Today I restored Greenspun links that were removed by anonymous (in a sense, editing without identification can be considered vandalism). jmswtlk 14:05, 22 March 2006 (UTC)

External links modified
Hello fellow Wikipedians,

I have just added archive links to 1 one external link on ICAD (software). Please take a moment to review my edit. If necessary, add after the link to keep me from modifying it. Alternatively, you can add to keep me off the page altogether. I made the following changes:
 * Added archive https://web.archive.org/20041122204622/http://www.ds-kti.com:80/our_products/icad.shtml to http://www.ds-kti.com/our_products/icad.shtml

When you have finished reviewing my changes, please set the checked parameter below to true to let others know.

Cheers.—cyberbot II  Talk to my owner :Online 13:41, 13 January 2016 (UTC)

External links modified
Hello fellow Wikipedians,

I have just added archive links to 3 one external links on ICAD (software). Please take a moment to review my edit. If necessary, add after the link to keep me from modifying it. Alternatively, you can add to keep me off the page altogether. I made the following changes:
 * Added archive http://web.archive.org/web/20110718093721/http://www.ds-kti.com/the_kti_vault/events.shtml to http://www.ds-kti.com/the_kti_vault/events.shtml#iiug
 * Added archive http://web.archive.org/web/20110612230403/http://www.ds-kti.com/the_kti_vault/2002_ppt/Product-Vision-and-Roadmap-Overview-for-IIUG-2002.ppt to http://www.ds-kti.com/the_kti_vault/2002_ppt/Product-Vision-and-Roadmap-Overview-for-IIUG-2002.ppt#348,1,Product
 * Added archive http://web.archive.org/web/20050205213112/http://ds-kti.com:80/latest_news/index.shtml to http://www.ds-kti.com/latest_news/index.shtml

When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at ).

Cheers.—cyberbot II  Talk to my owner :Online 15:34, 29 February 2016 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified 1 one external link on ICAD (software). Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Corrected formatting/usage for http://legacy.coe.org/newsnet/Oct05/index.cfm

When you have finished reviewing my changes, please set the checked parameter below to true or failed to let others know (documentation at ).

Cheers.—cyberbot II  Talk to my owner :Online 01:46, 1 May 2016 (UTC)

Real example needed
This article is written "in universe" and can only be parsed by someone who already knows what the various terms mean. Most of the article talks about various market forces and fails to talk about the program much at all.

There is a single example of an aircraft wing at the very bottom of the article, but it is largely indecipherable to me. From what I have been able to parse, it seems the system lets you enter parts as Lisp defuns? And then put them together into groups-of-groups with other defuns? Is that the basic idea?

And what is the purpose of this process? The article tells us where it's used, but fails to mention why it's used. Is it to make a list of parts? Or is there some sort of logic that can be applied?

Pinging as they appear to be the main editor. Maury Markowitz (talk) 20:27, 8 May 2017 (UTC)


 * Good points. For now, let me point to a bit on Quora about knowledge based engineering which is the general view. ICAD was one of the early approaches that proved to be effective for its time. ICAD started on the Lisp machine. Again, there was a query on Quora (What is a lisp machine and what is so grate about them). Common Lisp ported to the general Unix workstation. So, more work with ICAD was done under that environment. This work was under the general framework of CAE plus that deals with cradle-to-grave product management. In terms of real-world, I point to the Boeing 777 project which was the first large commercial plane that was digitally designed. The introduction of KBS (and ICAD) proved to be powerful for that program as it assisted in the integration of mathematical and other modeling results into a comprehensive view that spanned the design/develop cycles and beyond. That success got attention (oh yes, the magic of it all), so subsequent programs worked to include more (and more and more and ...) KBS. ... I'll see if this work has been written up; also, I will look for other published examples. Some links have gone stale in the last 1/2 decade, it seems. jmswtlk (talk) 00:51, 10 May 2017 (UTC)
 * Example citation to be found: "KBE, as implemented via ICAD, received a lot of attention due to the remarkable results that appeared to take little effort."
 * Suggestion: Engineering is a middle-out exercise. Take the Empennage, as an example. The design would be driven by several factors, such as mission, loads, looks, etc. That is, the old form versus function thing. KBE allowed a tighter cycle. It got to where a change to the structure by the performance modeling would be rerun over a weekend with the new geometry ready to go on Monday morning. Similarly for the rest of the structure. And, the skin? Bridging the mathematics and the operational aspects related to creating and forming the part was crucial. Hence, tooling and downstream processes made use of KBE, principally through the design-build focus. jmswtlk (talk) 10:56, 10 May 2017 (UTC)

So again I'm confronted with the problem of your description only being understandable to someone who is already familiar with the topic. KBE allowed a tighter cycle, how exactly. KBE provided a bridge between math and ops how. What exactly do ICAD actually do? Maury Markowitz (talk) 22:02, 10 May 2017 (UTC)
 * We would need examples that are public. Why? Much of this type of work is highly proprietary (hence cloaked). Engineering professors used to love to visit, of course, under information restraints. So, there are many example that I know personally.
 * Example 1: Let's take Rapid Forging Design (example of design rules). jmswtlk (talk) 13:57, 11 May 2017 (UTC)


 * Related story (The machines that made the Jet age). The technology is fairly complicated. The system (rules plus) allowed quicker definition of the die by the design expert.The die was created based upon the base part which conformed to the overall design of form and function (see Example #2). In terms of RFD, the part was defined in CATIA. The forging expert made a copy and introduced tags within 3D space that were marked on planes that depicted the shape. The markings identified the behavior of transition areas which were represented in fillets which joined in complicated manners.
 * The part model with its markups were passed through RFD (KBE) which interpreted the markings and built, piece wise, the model of the die. The die enveloped the part in a closed fashion that had a split to differentiate the upper and lower part. The forging process would insert the hot ingot into the lower part; the press would then come down with the upper part. The geometry needed to be closed; too, there were smoothness criteria that would prevent any type of constriction on the metal flow. In many cases, there were test tags introduced into the design at various points.
 * The forging process creates near-net parts. Final shaping processes remove excess material and smooth where necessary.
 * Recent example of near-net (image from Forging Magazine). This example is of a forging product that was converted to a process which uses 3D printing (related article -

Norsk Supplying FAA-Approved 3DP Ti Parts to Boeing).
 * Example 2: General frame assembly. To show this, let's look at a photo of the Fuselage structure. Let's bring over the image.There are lateral pieces (stringers - into and out of the photo). There are the vertical pieces (frames). There are ties that bind these two. All of these are constrained by geometry which is determined by program specifics. The stringers conform to (and hold) the skin.

Note: The photo is from a 747 which is two decades before the 777. In the earlier days, blueprints ruled. Trying to find symmetrical parts was a key item. In fact, the plane was a sequence of conical sections from nose to tail. With the advent of computer modeling, geometry (and the related topology) allowed freedom at the cost of complexity. With the 777 structure (find image of 777 structure or later - say, new generation 737), there was an added feature of the shear tie. These were specific to an area, however there was effort made to find commonality. With the 747, shims were still the norm. With the 777, shimming was much less as parts fit together. Several programs later (after the 777), there was sufficient modeling capability to think of pieces snapping together (or even being merged - by casting or forging or by use of composites). Take a shear tie (say), if it were to be solely designed by the conic approach, it would be easier to handle (yet, what variance would there be to the needed form). With modern modeling (b-spline), an additional step was required to reduce the representational footprint while maintaining form and fit (to a very small tolerance). jmswtlk (talk) 13:15, 12 May 2017 (UTC)
 * KBE created smart parts that allowed the frames to adjust to the skin with the stringers and ties conforming. So, any change to the skin (due to aerodynamic studies, shape, etc.) would require the frames, stringers and ties to adjust. On a huge plane, there are many of these. In the past, tricks like making the plane parts symmetrical (around a center line) helped. With the newer methods, demand for self-adjusting parts became apparent.
 * ICAD (essentially, a macro of a defpart entity) allowed this early on. It was flexible due to Lisp. Engineers were trained to specify inputs to the macro. Once defined and tested, the whole of the configuration could be rebuilt in the CAD system in a matter of hours and not days/weeks. Too, quality improved (put here link to studies of what the 777 brought forth in terms of make/manage).
 * All of the engineering was mathematically framed. Hence, automation of those steps could result in disparate modeling forms. Hence, representation changes were required. However, this type of work is not unlike what we see with refactoring of code. We had to refactor the model in order that downstream processes could use the results.


 * These examples are only two of very many related to aircraft. Not only were parts defined, tools were specified, and process steps were optimized, many times, used KBE (and related disciplines). In terms of these two, it must be recognized that the master data was handled by a data base. The dimensional view was managed by a major CAD system. In the time of the 777, auxiliary steps were outside of that framework. So, part of the work was extracting (checking out) data and returning updates. There were disparate systems (mainframe, workstation, PC).
 * Today, PLM is the driver that pulls together CAD, CAE, KBE and more, essentially CAX which has been written up here on Wikipedia.jmswtlk (talk) 18:33, 11 May 2017 (UTC)
 * Example 3. Changes in representation.
 * Each analytic method requires the model to support their reasoning (as in mathematical operations). The particular part will be the same as represented in the database and in the geometric configuration. Yet, there has to be a conversion in order to the the analysis. There are several types that are required. At the end, the results need to be put back into the database and CAD system. Too, downstream process, such as machining, may have their own requirements. Hence, these types of translations are commonly needed. Too, there needs to be concern that the modifications are true to the master definitions. KBE/ICAD proved to be quite useful in this type of work and helped prototype approaches that were less cumbersome. Along with the modification, analysis routines have complicated inputs. KBE/ICAD was useful in setting up parameters and other controls.

Real example, fuselage work
In the Reference section, #10, #11, and #12 have material that is specific to the 777.
 * 10 provides a Thesis in which analysis results are shown from the viewpoint of Management and Chemical Engineering. See Figure 6.7 for a view of the relationship between the skin, frame, stringer and shear tie for a panel (in this case, the Crown panel). The Thesis gives a good notion of the fit that was achieved. This analysis was several years after the initial design process. KBE was first introduced with the Program.
 * 11 was sponsored by Boeing. The presentation mentions the various types of computer-based analysis that were becoming possible. In the early stages of design, such work would produce results that had an impact on parts (general sense, particularly structural that were associated) that required adjustment of fit. A master dimension data base tracked these changes. Each discipline had to conform. A major CAD system provided the interface to all processes. KBE helped keep things in sync (such as adjusting geometry as required after analysis).
 * 12 deals composites, however they are compared to earlier approaches. Figures 15 and 16 deal with the 737, however it is from a Program that was after the 777 (and applied lessons learned). Figure 18 (see image this page - that needs to be replaced) is from the 747. Figure 22 has a distant shot of a fuselage under assembly for the 777. Details are more readily seen in the 737 shot. For the composite plane, 787, there is some discussion of shear ties.

We could find a lot more to support this section. jmswtlk (talk) 20:30, 24 May 2017 (UTC)

I just put a comment on the Talk page of the KBE article. I have the conference proceedings from a 1994 meeting in which lots of examples were given, some ICAD. One in particular dealt with tool design for an empennage rib. In the meantime, this is of interest, generally: Practical issues of AI (myself author). The TOC from the proceedings follows the paper in the file. jmswtlk (talk) 02:12, 15 January 2019 (UTC)

Earlier there was a link to a class at Stanford with wireframe view of the structure. That disappeared. ... Here is another example: Basic Aircraft Structures. Consider, that in the 777 work, KBE touched, at least, frames and stringers (and related parts) in about all the sections. jmswtlk (talk) 00:26, 12 November 2023 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified 2 external links on ICAD (software). Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added tag to http://www.lr.tudelft.nl/live/pagina.jsp?id=f1030549-cee5-4197-a4b0-2197fe671949&lang=en&binary=%2Fdoc%2FKBEtoFEmodelGen-Rocca.pdf
 * Added tag to http://www.lr.tudelft.nl/live/pagina.jsp?id=f1030549-cee5-4197-a4b0-2197fe671949&lang=en&binary=%2Fdoc%2F2002_5447-Gianfranco.pdf
 * Added archive https://web.archive.org/web/20060619194319/http://www.coe.org/newsnet/may03/icadnews.cfm to http://www.coe.org/newsnet/may03/icadnews.cfm
 * Added archive https://web.archive.org/web/20071026193646/http://www.coe.org/coldfusion/newsnet/june03/implementation.cfm to http://www.coe.org/coldfusion/newsnet/june03/implementation.cfm
 * Added tag to http://www.lr.tudelft.nl/live/pagina.jsp?id=b5411c30-06d3-424a-849b-085b774687c5&lang=en

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 12:22, 17 September 2017 (UTC)
 * Thank you. I will see what happened to Delft's material. jmswtlk (talk) 19:29, 17 September 2017 (UTC)

External links modified
Hello fellow Wikipedians,

I have just modified one external link on ICAD (software). Please take a moment to review my edit. If you have any questions, or need the bot to ignore the links, or the page altogether, please visit this simple FaQ for additional information. I made the following changes:
 * Added archive https://web.archive.org/web/20110304212021/http://eprints.rclis.org/handle/10760/11158 to http://eprints.rclis.org/handle/10760/11158

When you have finished reviewing my changes, you may follow the instructions on the template below to fix any issues with the URLs.

Cheers.— InternetArchiveBot  (Report bug) 01:39, 10 November 2017 (UTC)