Engineer to Order is a great way to increase sales and improve margins for some and a business reality for those in industries that demand products that fit into their customers’ environments. It’s also one of my favorite topics because I have seen the issue from the ERP / sales configurator side of things as well as the Engineering / CAD side of the business. I was really happy to have the opportunity to participate in an ETO seminar to discuss these issues with a group of mid-sized manufacturers along with my friends from Siemens PLM that work on their rules-based ETO technology Rulestream and Mercury PLM Services. I had the chance to talk to companies that produce electrical enclosures, custom windows, industrial washrooms, high end storage units, marine propulsion, firetrucks, and more. Two of my standard examples from past experience are custom windows (I make the argument that engineering custom windows is more complex then engineering a car, at least in some ways) and firetrucks (who knew that each one is different?). It was great to have a company that does each in the audience, I hope my examples were up to date (note to self, check in with them to refresh your examples!).
A Shift in ETO Challenges Toward Agility?
What are their biggest challenges? The overwhelming answer was time. It takes too long to develop accurate drawings and models for the shop floor. I also heard the most common issue that I hear in the ETO business, the ability to rapidly develop accurate quotes. Companies typically have to choose between developing quotes quickly (and overprice with a big cushion or risk losing money from underpricing) or developing them quickly. 
These issues are alive and well, but at least in this group the pressure of speed seemed to be on the rise. Maybe this is because of the instant gratification expectations we have today? Or because midsized manufacturers compete on agility and speed? I don’t know, it’s an interesting observation. It’s also a small sample size, so it may just have been the collection of companies that showed up for this seminar on a Thursday morning. It would be fun to do some more research on this to see if there has really been a shift.
How to Improve ETO Performance
I shared some of my research with the group that shows that top performing companies (those that have higher revenue and margin growth) are not more likely to engineer to order. I know, it wasn’t what they expected to hear. What I told them is that it isn’t the fact that they engineer to order that sets them apart, it’s how they do it. My research shows that top-performing companies are more likely to use rules-based design approaches that dramatically help improve quote accuracy and shorten order engineering time. The research also showed that top performing manufacturers are much more likely to use configuration / design automation technology. 
Some Great Examples
After I set the stage with some of the business implications and how manufacturers can improve ETO performance, I heard some good examples of this today from Mercury Marine and S&C Electric. S&C Electric has been using design automation technology for quite some time and shared how they use rules to eliminate waste and redundancy and reduce content creation time through automation and reducing defects. Carl Breving of S&C also talked about the importance of capturing engineering knowledge in their platform, another great benefit of design automation. They shared a common knowledge challenge where they have two people that engineer their orders that have combined experience of 90 years! Clearly important to capture that expertise! They shared a demonstration that showed how changes to specs were updated in the CAD model in what looks like real time. Pretty cool. 
Mercury Marine has products that span from build to stock to ETO. They use a 150% BOM with options and variants approach with PLM to drive configurations, which works well in many circumstances. They now have more complexity in a new product line that goes beyond this approach. One example is smart technology that includes more systems. They are looking toward more automated systems to help with the design – a great example of how rules-based technology can span beyond ETO. Another example was replacing a manual test in CAD that is required to meet a “tip test” regulation for fuel tanks, again an example of rules-based design automation that applies outside of the “to order” world.
Parting Thoughts
As a side note, I knew the Rulestream team and followed their technology prior to their incorporation into Siemens PLM. I have always been impressed with their ability to drive new CAD assemblies using parametrics based on rules. It’s a pretty cool approach and it was great to see their progress over the years. Brian Grogan of Siemens PLM also shared some additional examples and some very practical advice on how to implement rules-based design solutions based on a wealth of experience.
Rule based design automation has the basic assumption that design (Or I would say geometry) of the product will remain the same for significant amount of time. Enough time so that it would allow the product to be launched into the market and cross break even.
In my opinion , rule based design automation works best when parts/assemblies have LXBXH type dimension schema and relatively simple geometry.(or for the assemblies component placement)
Jim-
Thank you for your article. I was hoping that you would have explored the difference between ETO and CTO. It has been my experience that people confuse and interchange these terms.
At my workplace we run a rule-based automation system against multiple product lines to generate a MBOM, 3D models, flat pattern files (for sheet metal components) and two separate drawing packs- customer drawing and product pack. These drawing packs still require manual changes to correct or fix dimensions or design elements that haven’t made it back to the rules and master models for update.
We also run the same system for products that are either branded third party or manufactured but are not engineered. These products rely on selection rules but do not generate any data outside of identifying product numbers, costing and weight.
We have extended the product configurator to the sales team and struggle with getting the sales team to ask the appropriate questions of the customer to actually identify what product feature, functions, sizes and options they need to fulfill the architectural/engineering requirements.
Automation systems are great at generating data. But you can generate a lot of “bad” data quickly that will never be used because the sales team didn’t identify or the customer didn’t share all the requirements share up front.
Another issue is how to control changes in the rule-based/automation systems. It used to be that CAD models and the PDM system where the masters. With these systems CAD and PDM are just working components. The rules are the new master! Making changes to the rules and updating existing data to reflect these changes is a different animal altogether!
Ryan,
You bring up so many good points. Before I get started with the reply, I want to let you know that we’re running a survey on configurators (launched this coming week) that I would love to have you participate in. It’s a web-based survey. If you know others with the same ETO/CTO manufacturing challenges, it would be great for you to share it with them, too.
One of the key things I’m hearing from you is that your business faces different configuration challenges for different order / manufacturing scenarios. Many manufacturers find it hard to move design automation to the sales / quote process. The tools just aren’t made for that. What you probably want at that point is enough information to capture requirements, provide some validation so you don’t sell what you can’t deliver, and understand high level costs so you can price it. There’s probably a lot more detail needed to produce it (and maybe even to quote it, it sounds like your products are pretty heavily engineered to order specs). That’s where the design automation comes in. But sales (or even customers / channel partners in some industries) usually need to be taken through a more simplified process to just gather the need data you need to design it.
By the way, you aren’t alone in having to make changes to drawings, etc. after using design automation. Most companies can’t drive every decision based on rules. They typically try to automate the routine decisions and leave the more unique order requirements for an engineer to analyze and address. It’s often the typical 80-20 rule.
And about your comment on change controlling / versioning / lifecycle managing rules, all I can say is Amen! That is a very critical best practice.
Hopefully the report from the survey will give you more of what you hoped to get from the original article you replied to. Good timing!
Thanks,
Jim
Jim-
I was hoping to get more out of this article 🙂 . Starting with a definition of rule-based capture engine vs. a straight out product configurator would have been nice. Understanding that we are talking about different base components helps the reader understand that they are not tied to one specific product suite for design automation.
Capturing your design/engineering rules in at top level software (independent of your CAD), to me, is a huge strategic decision and advantage. Having a tool that can pass on your same set of design rules to multiple CAD systems provides a company the flexibility when looking at M&As. Knowing you are not tied to one specific CAD system (i.e. DriveWorks and SolidWorks partnership) allows you to expand your M&A options and reduce the amount of CAD dependencies.
Ryan,
I apologize, I missed your second post on this. I’m going to take your input and channel it into a new report I’m writing. The survey is live on my site now, feel free to share it with any of your ETO/MTO manufacturing friends. You can access the survey directly at https://www.surveymonkey.com/collect/?sm=q94yeBg98M2yPmB4BzO3CaU54h8PWRTAT6WDU_2B_2FloLa3PBBqk7PrQnBPZAoCzWbW. One of the key things we are looking at is the role of Sales Configurators versus Technical / Engineering Configurators.
I agree on your view of managing rules, although most of the implementations I see are tool-dependent (maybe not CAD tool, but configurator tool specific).To add to that, they should be managed properly (revisions, change management, etc.). They are a tremendously valuable asset!
Best,
Jim