Design Automation Using Machine Learning & Computational Design to Make Products Manufacturable
With Rhushik Matroja of CDS (Cognitive Design Systems)
Co-founder and CEO Rhushik Matroja saw that problems evaluating and optimizing designs for additive manufacturing could be accelerated using machine learning and computational geometry synthesis in an efficient and scalable manner so expanded the team and software offerings to support not just optimization for additive, but also other manufacturing processes such as casting, machining and molding.
As we get closer to the CDFAM Symposium where Rhushik will be presenting, we asked him a few questions about CDS, their software applications, use of AI and how it can help overcome the engineering and business obstacles in adopting advanced manufacturing.
Can you start by giving us a brief overview of your background and the experiences that inspired you to establish Cognitive Design Systems as a company?
Through our extensive collaborations with prominent automotive and aerospace companies, we have acquired profound insights into the transformative nature of design concepts over time. The journey of a concept involves a series of critical stages, including CAD generation, CAE validation, design reviews, configuration management, manufacturability analysis, and cost analysis.
By integrating these essential KPIs into the purview of design engineers, we recognize the pivotal role they play in the development of successful products.
Our ultimate goal is to automate and expedite the product design process, empowering design engineers to transcend their role as mere button-pushers and become decisive decision-makers. Essentially, we aim to automate the tasks that Henri and I, as Design and Manufacturing consultants, have been undertaking.
Our founding team comprises Henri de Charnacé (CTO), Vincent Ung (COO), and me, Rhushik Matroja (CEO).
Both Henri and I bring expertise in mechanical engineering, while Vincent’s background lies in finance. Our team is composed of highly qualified engineers and programmers hailing from esteemed engineering institutions in France.
With a diverse array of talents encompassing material research, mechanical engineering, design and simulation, software engineering, and even an ex-banker, we are well-equipped to tackle the multifaceted challenges in our field. Headquartered in Toulouse, France, a city renowned as the European capital of the aeronautics and space industry, we are strategically positioned to contribute to this thriving domain.
Can you describe the core components that define your software’s capabilities, and provide an overview of some applications or use cases where your software is commonly utilized in real-world scenarios?
Cognitive Design Systems is at the forefront of developing a diverse array of software and technologies catering to multiple manufacturing processes. Presently, our commercially available software, Cognitive Additive, focuses on Additive Manufacturing. However, we are pleased to announce that in the summer of 2023, we will be introducing Cognitive Molding, a specialized software dedicated to Injection Plastic and Aluminum Die Casting.
The foundation of this software is built upon two core technologies, namely:
1. Manufacturability analysis
2. Advanced and robust geometrical engine
Leveraging the capabilities of these two foundational technologies, we can generate designs that are both manufacturable and marketable.
Our assessment process encompasses evaluating the manufacturability, cost implications, and carbon footprint associated with each part. Through this comprehensive assessment, valuable insights are gained, facilitating the optimization of the design. To achieve this optimization, we employ a range of established technologies such as topology optimization and parametric optimization, ensuring the automation of the design enhancement process.
Within the realm of large original equipment manufacturers (OEMs), project managers often encounter considerable challenges when endeavoring to justify the suitability of a particular part for Additive Manufacturing (AM).
To address this predicament, CDS has developed an exemplary part screening software, denoted as Cognitive Additive, which possesses exceptional capabilities in analyzing vast quantities of parts according to diverse criteria encompassing manufacturability, serial production cost, carbon footprint, and other relevant factors.
The comprehensive analysis undertaken by this software incorporates an AS9100-inspired costing model, which diligently accounts for various expenses associated with AM setup, AM process, post-processing, inspection, and testing. Moreover, it effectively orients and nests the parts in either a 2D or 3D context, thereby providing a holistic understanding of cost and manufacturability.
This analytical process gains significant acceleration through the utilization of machine learning algorithms, enabling precise and on-premises computations. By leveraging a substantial database of 3D models or assemblies, the software facilitates the identification of optimal manufacturing processes, suitable machinery, and potential candidates for part consolidation.
Utilizing our manufacturability analysis engine and geometric engine, our forthcoming software exhibits the capability to identify and selectively alter geometries on a local scale, thereby enhancing manufacturability.
Automated design for casting stands as a prime illustration of this technological prowess. Through the implementation of Cognitive Molding technology, automotive original equipment manufacturers (OEMs) can seamlessly transform their conceptual designs into die-castable components.
The progression involves the creation of molds and subsequent optimization for manufacturability via additive manufacturing (AM) or milling, entailing modifications to the initial geometry. This iterative procedure gains acceleration from the utilization of our formidable geometric engine.
Optimization of Topology, Manufacturability, and Cost:
In collaboration with Synera, Cognitive Design Systems offers valuable assistance to our shared clientele in optimizing components based on three out of four key performance indicators (KPIs) crucial for achieving exceptional product design.
This pioneering multi-parameter optimization workflow stands unrivaled within the industry, enabling the identification of optimal trade-offs between performance, manufacturability, and cost.
Consequently, this workflow significantly reduces design cycle time, leading to enhanced efficiency. For further exploration, we encourage you to explore the Cognitive Additive Plugin for Synera.
Could you give us a brief overview of your upcoming presentation at the CDFAM symposium on “Design Automation Using Machine Learning & Computational Design to Make Products Manufacturable,” and how it relates to the features and applications of CDS software?
We leverage our advanced manufacturability analysis algorithms in conjunction with our proprietary hybrid geometrical engine to transform designs into a manufacturable state. Specifically, for die-casting applications, our algorithms automatically modify non-manufacturable zones and incorporate draft angles to enhance the manufacturability of the design.
In the realm of additive manufacturing, our software intelligently adjusts the part geometry based on Design for Additive Manufacturing (DFAM) principles, addressing considerations such as overhangs, bridges, holes, and thin walls.
By gaining insights from the manufacturing stage, which represents the final phase of the design chain, we establish a closed-loop system that allows us to iterate through a vast number of generated designs. This iterative process enables us to identify the ideal part solution for the user.
It is important to note that there is not a singular solution for generated designs and topology optimization; our objective is to determine the best-fit solution among the multitude of possibilities.
During our presentations, we showcase the thoroughness of our manufacturability analysis and demonstrate the capabilities of our geometric engine in effectively enhancing the manufacturability of parts while adding significant value to the overall design.
With the current hype and misinformation surrounding AI, particularly in the engineering field, could you clarify how CDS is effectively utilizing AI in the processes of design, optimization, and cost analysis within engineering and manufacturing sectors?
We harness the power of artificial intelligence (AI) for several purposes, including:
1. Accelerating time-consuming operations: Through the utilization of AI, we can expedite operations that are traditionally time-consuming, such as calculating printing times for additive manufacturing and performing manufacturing risk analysis. By leveraging AI algorithms, we streamline these processes and improve overall operational efficiency.
2. CAD and mesh geometry and data enhancement through Graph Convolutional Networks (GCNs): By employing GCNs, we can modify mesh and CAD data, facilitating seamless integration between different design stages and enhancing the overall design workflow.
Additionally, we are actively working on other confidential projects that we intend to patent in the future. These developments encompass a range of cutting-edge technologies and methodologies aimed at further advancing our capabilities in the field of design and manufacturing.
How do you collect and manage data for training these algorithms? Is there a shared database for all clients, or do you offer the option for each client to create a segregated data repository as a ‘walled garden’ to prevent their information from benefiting competitors?
At present, we do not collect data directly from regular customers due to confidentiality and security considerations in the aerospace and defense sectors. Instead, our solutions are developed locally using a combination of open-source 3D libraries, our historical data, and data obtained from trusted partners.
This approach allows us to generate the necessary data to train and optimize our algorithms while ensuring the protection of sensitive information. By leveraging these diverse data sources, we can develop robust and effective solutions tailored to the specific needs of our clients in the aerospace and defense industries.
What are the primary indicators that suggest a company should consider exploring CDS software to address their challenges, and is it necessary for them to provide additional data beyond basic geometry for effective problem-solving?
1. Cognitive Additive is the ideal solution for companies facing challenges in identifying parts suitable for Additive Manufacturing (AM) serial production.
Our software efficiently evaluates thousands of parts for manufacturability, cost, and other critical criteria, empowering companies to optimize their AM processes and make informed decisions. With Cognitive Additive, companies can unlock the potential of AM by identifying the right parts for serial production.
2. Design Optimization Needs: Companies seeking to optimize their designs for performance, manufacturability, cost, and carbon footprint can benefit from CDS software. Cognitive Additive’s advanced algorithms and capabilities enable design engineers to make informed decisions and achieve optimal design solutions
3. Additive Manufacturing and Casting Applications: Companies involved in additive manufacturing or casting processes can greatly benefit from CDS software.
The software’s ability to automatically modify designs based on DFM rules, remove non-manufacturable zones, add draft angles, and optimize part geometries specifically for these processes ensures enhanced manufacturability and successful production.
4. Custom Applications: In addition to our core technologies, we offer stand-alone software and Software Development Kits (SDKs) that enable the creation of custom applications.
Our SDKs provide a flexible framework for developing customer-specific applications tailored to unique industry needs.
We have successfully developed applications for a range of industries including MRO in the defense sector, footwear, sports equipment, and the medical industry. These custom applications leverage our advanced technologies to address specific challenges and deliver tailored solutions to our clients.
What do you perceive as the most common barriers to adopting this technology, and how does your company assist clients in overcoming these obstacles?
A significant challenge in the design software industry is achieving digital continuity, particularly when it comes to integration with specific Product Lifecycle Management (PLM) and Manufacturing Execution System (MES) platforms used by major Original Equipment Manufacturers (OEMs). Preserving the integrity of the data chain is crucial, and breaking it remains a sensitive subject.
To effectively reach a broader audience, collaboration, and integration with these platforms are essential for the success of Cognitive Design Systems (CDS). As part of our strategy, we have formed partnerships with various large organizations to bring our technology directly to customers.
Our collaboration with industry leaders like Stratasys and emerging engineering platforms such as Synera exemplifies our commitment to a collaborative approach. Through these partnerships, we aim to integrate our technologies seamlessly into existing PLM and MES platforms, enabling a smooth and unified workflow for users. This collaborative strategy allows us to expand our reach and deliver our innovative solutions to a larger customer base.
By working closely with established organizations and integrating our technologies onto widely used platforms, CDS is dedicated to overcoming barriers and ensuring the successful adoption of our advanced design software in the industry.
The additive manufacturing industry, being a rapidly expanding sector, must educate a broader audience to enhance technology adoption. By providing decision-makers with manufacturing and business insights, we enable them to adopt additive manufacturing on a larger scale, finding applications that align with business objectives while delivering optimal performance. At CDS, our focus extends beyond additive manufacturing, as we aim to provide users with insights into various manufacturing processes. To ensure the right manufacturing process is connected with part design, we offer valuable insights on feasibility, cost, performance, and other key performance indicators (KPIs).
Lastly, what are the key takeaways you hope attendees will learn from your presentation at CDFAM, and what do you personally aim to achieve or gain from participating in the event?
As part of our mission to collaborate with larger companies and expand our reach, we actively seek synergies among speakers and audiences. We recognize the value of connecting with like-minded individuals and organizations who share our vision for innovation and technological advancement.
During our presentations and engagements, we are eager to provide detailed information about our software solutions, Cognitive Additive and Cognitive Molding, to interested participants. We understand the importance of showcasing the capabilities and benefits of our technology, and we strive to offer comprehensive insights to those who are curious and eager to learn more.
By fostering these connections and sharing knowledge about our software, we aim to create meaningful partnerships and opportunities for collaboration. We believe that through these engagements, we can further enhance the understanding and adoption of our innovative solutions, ultimately driving advancements in the fields of Additive Manufacturing and Molding.
We look forward to engaging with participants, sharing information, and exploring synergies that can propel our technology to larger platforms and make a positive impact on the industry.
Rhushik Matroja will be presenting alongside other experts in advanced computational design and engineering, AI and advanced manufacturing at CDFAM 23 in NYC.