Discover more from Bits to Atoms
DfAM Q&A With Rhushik Matroja of Cognitive Design Systems
"Our Goal is to Automate the Design for Manufacturing Process"
Our first Q&A with DfAM consultants in this series is with Rhushik Matroja of Cognitive Design Systems. While Cognitive is first and foremost a software company, their software, Cognitive Additive is entirely informed by the DfAM consultancy work they have collectively executed in the past.
What led you to form Cognitive Design Systems, who are the team members and what is their background.
While working with some of the biggest companies in automotive and aerospace, we came to know how the concept conceived by the designer changes over time. The concept goes through CAD generation, CAE validation, design reviews, configuration management, manufacturability analysis and cost analysis. By the time it is ready to manufacture, the essence of the product is completely modified. With my friend and now the CTO of CDS: Henri de Charnacé, I wanted to change it! So we have created software for DfAM called Cognitive Additive.
Our goal is to automate the Design for Manufacturing process.
Henri and I are mechanical engineers and Vincent comes from a background of finance. Our team is constituted of highly qualified engineers and programmers from the best engineering schools of France.
We have material research engineers, mechanical engineers, design & simulation engineers, software engineers and also one ex-banker! :) Our HQ is based in the city of Toulouse, France which is also the European capital of the aeronautics and space industry.
What are some of your favorite projects you can talk about?
Cognitive Design Systems is one year old, but we have already worked on more than 10 projects.
We have executed projects in aerospace, automotive, sportswear and the medical industry. One of the challenging projects was with a supplier of Boeing. They challenged us to find suitable parts for additive manufacturing from a subassembly of more than 900 parts. Using Cognitive Additive, we analyzed 900 parts by going through 61,000 combinations of various machines and materials in under 7 hours. Each part was analyzed for manufacturability, cost, complexity and buy to fly ratio.
We extracted top 30 candidates (originally made by milling) where the AM process can bring a real added value. On top of this we have identified 16 static assemblies which could be unified into a single part, as well as 26 parts with a little touch of DfAM which could become a great candidates for additive.
Usually such work in the past was taking 3 to 4 months of engineering work. Now it is just a matter of hours.
How does your company work, can you walk us through a ‘typical’ client interaction, and what is your business model?
We are foremost a software company. Our consultancy projects are extensions of our software capabilities. We provide consultancy on design for manufacturing with a special attention to Design for Additive Manufacturing.
DfAM as it is known, is not just a bunch of thumb rules defining how we should design a part. While designing a part we have to take into account manufacturing risk analysis, costs related to the quantities, post processes and of course thermo-mechanical performance of the part during manufacturing and it’s lifetime.
At CDS, we find solutions to engineering challenges using various softwares. Our design process is done using nTopology ( implicit modeling engine), simulations are done on Ansys and DfAM analysis is done on Cognitive Additive.
We also develop and use custom applications for design of experiments and manufacturing optimization. In the following image, we can see a typical consultancy project.
What is the business model for your software, is it on premise or cloud based and what is the ramp up time required to be able to use the software?
A. Cognitive Additive is an on premise software with an annual subscription based business model. There is no cloud connectivity required so you can use it off grid.
This was one of the most important requirements from our early adopters in Japan. However, we will bring a cloud hosted software later during the year in order to introduce our services to a larger public.
There are three main user profiles for our software.
AM service bureau sales rep
Cognitive Additive is primarily used by the design engineering team working on new product development. Designers and engineers can understand the impact on cost and manufacturability at each design iteration using our software. We are working on digital continuity by creating apps for CAD softwares.
Second type of users are service bureau sales reps. Only one hour of training is required before becoming completely autonomous on the software.
The third type of user profile is project manager. While making decisions on the manufacturing process of each part, and doing economical viability study on each part of the product, Cognitive Additive provides them with the right KPIs to compare.
Given your company is focused on design for additive manufacturing, how do your clients know that an AM design is the answer to their engineering requirements?
Well, additive manufacturing is NOT the solution to everything. First step is to find the right manufacturing process based on performance, cost and productivity.
Our software, Cognitive Additive is tackling this very issue. By pulling the data on performance, cost and manufacturability of the part, so we can understand whether AM is the solution or not.
Second step is to improve the manufacturability and cost of the part. We call it ‘manufacturing optimization’. Currently this step is proposed in our consultancy service, but we are working on automating it using machine learning tools. Keep your ears open for it!
Once you have a design/solution that meets the engineering requirements, how do you also ensure the solution meets the business case?
This is a very interesting question! Basically, we ask Cognitive Additive for the answer. We can export a dashboard for each part like the one below:
On the right side, we have two graphics related to the cost analysis. Above is the evolution of the part cost using the given machine/material/parameter set, and below is the cost distribution while manufacturing. Understanding the real cost of the part could be very tricky, as it involves almost 50 parameters. You can see what these parameters are in the chart below:
What are common misunderstandings clients have of DfAM and additive manufacturing in general on initial engagement?
There is a lot of “Fake News” circulating in the industry about additive manufacturing. Here are few of them:
Additive is for prototyping,
Additive can’t hold high stress environment
Additive is not accurate
Additive is too expensive
Additive is not qualifiable with standards
Additive is not ready for industrial challenges
We use the real case studies from the industry leaders to show that additive can be a viable option. In other words, we provide them with our software Cognitive Additive, to do ‘fake news busting’ with some real data!
What are some issues you see arise that block the adoption of additive manufacturing as a solution.
Finding the right application for additive manufacturing is a highly underrated issue in the industry. Additive is regrouping a very large scope of manufacturing processes. Comparing apples to apples is important. I would make jigs and fixtures in FDM but wearables in MJF or DLP not the other way around. Cognitive Additive will be able to tell you exactly this in an autonomous manner.
Design for additive manufacturing is a lot talked about but not enough practiced. Manufacturing a part designed for aluminum sheet metal, by Powder Bed Fusion is not fine! It can only give you a bad reputation for the technology.
Manufacturing Cost! After all, we need to work hard to bring down machine, material and post process costs. A better equipped and trained ecosystem can make it happen. Only time will tell us whether it is possible to have it in every corner of the world.
What most surprises you about clients requests or expectations?
Their budget! :)
The design tools, additive manufacturing materials and processes are evolving relatively quickly, how do you stay up to date with the latest offerings?
Very interesting question.
New materials and manufacturing processes are entering the market at an unprecedented rate. Modern software is updating them every other week with new functionalities.
As a small company we have some highly passionate engineers who track the evolution and keep each other informed with the news.
Our ambition for Cognitive Additive is to integrate these new technologies as fast as possible. If you put yourself at the place of giant corporations like Airbus or Boeing, the cost of training hundreds of design engineers every year on new manufacturing technology is very high. It is easier to just update our software and bring the latest technologies to their PC.
Finally, what most excites you about designing for additive manufacturing and what are you looking forward to in the future.
In the past, my designs were hard to create on regular CAD software and even harder to manufacture them. But with the tools we have developed and what software such as nTopology offers, I am able to convert my concepts into a product without much hassle. Once the concept is ready, Additive Manufacturing gives me enough freedom to manufacture it.
I am looking forward to being able to design products more efficiently using AI!
We are working on it!
I would like to thank Rhushik for his time to answer our DfAM questions, we will follow up to do a more in depth review of Cognitive Additive software in the near future.