Speak to any harness manufacturer, and they will attest to one thing. Drawings that come from customer can be wrought with mistakes. The type of mistakes runs the gamut from incorrect components selection, mismatched contacts and wire sizes, routing issues, labeling errors, incorrect bundle diameters, length issues, and just about anything else you can think of.
There are a plethora of reasons these mistakes happen. Overall, the OEMs are resource poor in engineering. They have other systems to think about, and the harness is the last thing in any development process, and early to be installed. The sheer number of components is overwhelming, and of course the harnesses require the assembly of parts from multiple component manufacturers, so choosing the right ones can be daunting, especially for less experienced engineers.
Often times, new designs from an OEM may just be new iterations of an existing design, with any mistakes being carried over. Some customers may have a database of components that is dated or simply incorrect. OEMs lack the awareness and resources to update their database information, yet this is the list they rely on when designing a new harness assembly. Connector manufacturer delimiters on part numbers may change when a company changes hands or a product line is updated, but these old part numbers remain in thousands of libraries.
Some engineers and designers of the OEMs may pick components simply because of the component suppliers they have access to or the standards libraries within the developed database. They may have been shown ‘the latest and greatest’ and are eager to spec it in their next design without understanding compatibility issues. The component logic of new products may not be built into the CAD provider’s design rule check (DRC).
As there are a relatively small group of engineering and design firms that specialize in interconnect systems, and there is a lack of time and expertise to design internally, OEMs often resort to design by trial and error, which can be expensive and time-consuming and may ultimately result in an inferior product. Further exasperating this is the fact that new engineers are taught to design in space and encouraged to let the suppliers work out the details.
It usually comes down to the lack of time, expertise, and training on behalf of the OEMs. A great deal of tribal knowledge has disappeared as veteran designers retired without much thought to apprentice programs for new engineers. Perhaps harness manufacturers enable this behavior as they beef up their engineering staff.
The bottom line is it all costs money for harness manufacturers, usually in engineering time and subsequent delays. This is especially true with today’s supply chain constraints, where components have long lead times. How does a harness manufacturer scale up when constrained ironing out all these details?
Arik Vrobel is working on a solution. Having grown up in the harness industry, he has the experience to understand the problem. Until its recent sale to Winchester Interconnect, El-Com Systems was his family business, with close to 500 employees at four locations and yearly sales of approximately $50M. He has seen first-hand how constricting and expensive this issue is, and design validation has long been a pet project for him. Arik evaluated many existing CAD & engineering tools and found them lacking as a DRC in some key metrics.
Having completed the sale of his business late last year, he has embarked on a mission to solve this issue for the industry by creating his new company called Cableteque, which is in the advanced stages of developing a design validation tool they call Predictive Interconnect Analytics (PIA). PIA aims to enhance the design process by identifying potential errors and providing interactive feedback to engineers at the design stage. It is designed to operate alongside customary design tools and will work with any major CAD/ECAD system and non-CAD designs. The tool uses advanced machine learning algorithms, AI feedback, and built-in logic to “fill in” design gaps, such as the interconnectability of components or placements and sizing of consumables.
Cableteque’s Predictive Interconnect Analytics (PIA) software is a cloud-based AI solution that allows data synchronization with multiple databases (internal and external) options that will facilitate proprietary DRCs to minimize costly mistakes in the early stages of the product modeling & design, thus filling the gap in form, fit and function (FFF) requirements. This is done thru extensive algorithms of the DRC that compare the allowable interface values and features of the harness components from an extensive associated parts library’s technical data. It provides two levels of alert, either an error or warning, with a later feature of providing design recommendations and improvements.
Cableteque’s PIA software is an add-on tool to the designer’s CAD. It will integrate seamlessly with numerous native CADx tools by leveraging standard outputs of design tools such as the parts list, From-To lists, and other design data and provide comprehensive feedback so that the designer can target “design risks” to minimize mistakes and enhance confidence in the expected outcome.
The PIA solution is particularly valuable for high-reliability industries where electrical interconnects are highly complex and sensitive in terms of performance due to application parameters and constraints, which is further challenged by the implementation of rapid design changes.
Cableteque will be hosting a technical conference at the EWPT Expo in Milwaukee on May 18 at 10:00 am titled Predictive Interconnect Analytics – Enabling Your Design Potential. They will highlight known problems in the industry and ways PIA can solve this industry-wide problem. If you are unable to attend, take a look at the blog at https://cableteque.com