Hardcore Engineering
You can only admire the highly effective, lean engineering teams found in places like Lockheed's Skunkworks (back in the days) and SpaceX. These teams benefit from streamlined management structures, empowering smaller, self-sufficient groups with considerable responsibility and autonomy. Such an approach promotes direct communication and efficient sharing of knowledge, placing everyone in close collaboration. But making it work in reality is very difficult because it depends on having a highly skilled engineering team organized in an optimal way. Otherwise, you might take on more risk than you can manage, and things fall apart quickly.
Prototypes are Easy, Production is Hard
Perhaps Elon Musk’s most quoted insight from his Tesla experience is that production is about 100 times more difficult than building a prototype. Developing a hardware product and production system for high-volume production is significantly harder than creating a single, sometimes handmade, prototype. And, if the product is both high-volume and complex, the difficulty level increases dramatically. The truth is, those high-achieving engineering teams at Lockheed and SpaceX accomplished a lot with a small number of people, but focused mainly on prototypes and low-volume products. In today’s complex engineering landscape, where the number of machines & robots per human is significantly higher than in the past, a large portion of the global engineering resources is dedicated to production-related activities.
What is a Prototype?
Creating a prototype involves designing and building a preliminary version of a product. The goal of a prototype is to test ideas, features, and the feasibility of the design. Since the focus is on exploration and demonstration, there are fewer constraints. You don't have to worry as much about the cost of materials, the efficiency of manufacturing processes, or the scalability of production. You still need to design the product with those considerations in mind, but early prototypes typically give you some slack. This doesn't mean that creating a prototype is without challenges, but compared to mass production, it allows for more flexibility and experimentation without the pressure of scalability, cost minimization, and market demands.
Production is a Whole Other Beast
Transitioning from a prototype to mass production involves many additional considerations and challenges. These include ensuring the product can be manufactured reliably at scale, sourcing materials cost-effectively, setting up production lines, ensuring quality control, and addressing logistic challenges. Production also involves significant financial investment in machinery, labor, and materials. The complexity increases with the need to optimize production processes for efficiency, manage supply chains, and comply with regulatory standards. All these factors make production significantly more challenging than developing a prototype. And you need to iron out all the significant risks in the product design before starting production; otherwise, you could face significant challenges.
Failing with Hardware is Costly
Hardware failures are expensive, especially true for prototypes, but failures in production can have catastrophic consequences. The entire production system and supply chain, that depends on a specific product version, can be very costly to change after production starts. For prototypes, failure is somewhat expected to test performance limits, but unnecessary failures should be avoided. Ordering parts for prototypes can take weeks, and engineers often find that it is difficult to allocate this time efficiently, as they are ”waiting” to do prototype tests before continuing. If prototypes are planned sloppy, without a clear risk assessment (aka FMEA), you could end up testing unnecessary things and not address significant product risks.
Why is the Development Cycle So Long?
To avoid significant risks in production, it's crucial to ensure that the product entering production is robust. This process consumes the majority of engineering resources in product development, transitioning from a concept or early prototype to a product ready for production. It involves multiple cycles of prototyping and design iterations, and a lot of problem solving along the way. And in parallel to this, you need to set up the production system and supply chain, while coordinating sales and marketing efforts to generate customer interest and pre-orders. This entire process requires complex coordination of many moving parts in the organization, spans multiple years and demands substantial capital and resources.
What is Consuming Time in Development?
Ok, so you might think development engineers spend all their time on new innovations and creative design work, right? Well, they don't. Instead, the majority of their time is dedicated to generating repetitive designs in CAD, producing drawings, handling other documentation tasks, and searching for necessary information. The design and documentation process tends to be highly manual, and not all CAD work involves creativity. Many components in new products have been designed previously in some product somewhere, allowing engineers to utilize standard libraries for common parts like screws and bolts. However, when a slightly customized part is required, they typically must design it from scratch. Creating and maintaining drawings is perhaps the most common pain points since it consumes a lot of time but doesn't add any value to the product—it's just a description of how to manufacture it. Moreover, engineering tools often fail to provide guidance on utilizing industry best practices and creating effective designs, leading to unnecessary design iterations, chasing information, etc. Therefore, the reality for most engineers involves little time for creative work and a lot of context switching.
What is Needed? (Psst, keep this part a secret)
For us at Visendi, the solution is clear: engineers need a Copilot that eliminates non-creative and repetitive tasks. Since hardware engineers primarily work with CAD tools, this Copilot must integrate seamlessly into that environment and be able to interpret and produce CAD models. And this is what we're building at Visendi. In the short term, this Copilot will act as a smart extension of the CAD system, assisting with repetitive design tasks, offering guidance, and performing feasibility checks. But in the long term, we're aiming to build a digital design engineer—a true game-changer expected to boost productivity by at least 10x. This is how we plan to achieve a new wave of innovation within hardware, and faster get to milestones like establishing a base on the moon and a city on Mars. But first, however, we need to teach it to use engineering tools effectively....
Comments