1.) What MCAD software does Iced*CAD support?

Iced*CAD, today, is fully integrated with PTC Pro/E, Autodesk Inventor, and Solidworks. Iced*CAD works with any MCAD program with public API’s. While Iced*CAD is tightly integrated with Pro/E, Autodesk Inventor, and Solidworks programs, Iced*CAD can work with any CAD program’s public API’s and is independent of the underlying CAD program, so Iced*CAD should have no trouble working with the majority of CAD programs in use today.

2.) Your brochure mentions extracting data from assembly bills of material residing in a manufacturing system database. With which databases specifically can Iced*CAD communicate? How?

Iced*CAD can communicate with most ERP (Enterprise Resource Planning) systems, such as Oracle, JDE, SAP, or Peoplesoft, through standard SQL (Structured Query Language) queries. Integration with an industry standard database does not present a problem. Iced*CAD executes a SQL query to the backend system requesting the Bill of Material data for a specific known-good "part number" generated by Iced*CAD. The backend system returns that result to Iced*CAD which now knows all the 3D parts that it must “pull” from the clients PDM (Parts Data Management) or CAD system to create that assembly.

Part of the Iced*CAD implementation process is to make sure that a customers’ part numbering schema is valid and accurate, or if it is not accurate, to "clean-up" the part numbering system so only appropriate bills of material can be pulled from the system in the first place. This can be resolved in Iced*CAD or the customers’ backend system, whichever our customer prefers.

3.) What modeling prerequisites are there to ensure that Iced*CAD assembles components correctly, and doesn't jumble them up? How can our designers be sure that they do not need to go over an Iced*CAD assembly with a fine-toothed comb to prove it is correct?

The modeling prerequisites and engineering algorithms Iced*CAD uses are patented and proprietary, and are fine tuned to each customer’s needs. Iced*CAD will attempt to assemble any bill of material presented to it based on those rules and algorithms. However, if the assembly is not valid per the rules and algorithms, the parts of the assembly that did not "fit" are simply included in the drawing but left to the side as "extra" parts. So it is possible to assemble, say 90%, of the drawing and then have parts Iced*CAD didn’t know what to do with left over. Iced*CAD will attempt to build a system based on the "known-good" part number submitted to it.

We encourage the client to have an engineer review the generated assemblies for accuracy when the system is first launched. Mistakes do happen in the testing and roll-out of Iced*CAD but once Iced*CAD knows how to assemble a certain grouping of assemblies or sub-assemblies it should not make any mistakes on those assemblies again, no matter where it appears in any subsequent super-assemblies, nor if any new parts are introduced into the assembly. This is valuable as most systems are made up of sub-assemblies and each sub-assembly can be tested and perfected. So when those sub-assemblies are put together in the final assembly the engineer should have high confidence that all sub-assemblies and the final assembly are correct. This also allows engineering to create batch processes of assemblies. Thousands of 3D drawing variations of a certain model can be created for example and stored for immediate retrieval by a customer via a simple website. If any sub-parts change at the engineering level, Iced*CAD would recognize this and simply create a new 3D drawing of the assembly on-the-fly at the next request.

4.) Your literature mentions a high-end server is recommended, why?

Iced*CAD is server intensive. While a high-end processor is unnecessary, we recommend the most advanced server available to speed assembly. Typical assemblies can take anywhere from 3 to 10 minutes to assemble depending on the complexity of the calculations and the volume of parts that must be assembled.

5.) Our product can be reasonably complex (say 300-400 components per assembly). The product has many similar-looking components, some of which are flexible (springs, snap-rings, O-rings, rotary seals), and some of which are assembled in a specific way not necessarily handled by the CAD system (i.e. meshing gears). How has Iced*CAD proved itself with products of this complexity (any case studies?). How long might Iced*CAD take to automatically assemble a product like this?

Iced*CAD has proved itself in this area. One client assembles cartridges and manifold assemblies on the fly. There are over 25 million different combinations, although the individual bills of material are on the order of 20-30 parts per assembly.

Iced*CAD creates assemblies from parts but also creates assemblies from sub-assemblies. So the complexity is minimized. If the 300-400 components have to be assembled all at once, instantly, Iced*CAD could do it but the time to assemble could become prohibitive. However, it is important to note that Iced*CAD is always dramatically faster than a manual engineering process. As an example, if those 300-400 components are actually 10-20 sub-assemblies made up of 20-25 parts, Iced*CAD would assemble all the sub-assemblies first, and then assemble the final assembly. We have not come across a product line that could not be broken down into simpler, digestable, sub-assemblies.

6.) Does Iced*CAD replace my engineering staff?

No. Iced*CAD is meant to increase the efficiency and effectiveness of the engineering department not replace it. One study by EDS placed tedious, non-value add engineering tasks, such as assembly drawing creation, at 66% of an engineer’s day. At one client, they reduced assembly generation turnaround time to the client from two-weeks to a few minutes. Their Engineers now focus their energy on creating new products and innovating new competitive advantages rather than tedious assembly generation.

7.) What does Iced*CAD cost and how long does it take to implement?

Iced*CAD installations vary widely by assembly complexity and system integration requirements. Typical complex installations require 4-6 months and cost anywhere from $100,000 to $400,000+. The return on investment is usually, however, very high, as our customers are able to impact quoting, sales, engineering time, customer service, shop floor assembly time, quality, and production and shipping. A payback of less than 12 months would be a realistic target and is documentable.