Working in flight test is a unique experience, especially for a human factors engineer. Having access to highly trained and specialized users increases the scope for what is possible in a system. The training and certification process involved in becoming a pilot means that the aircraft doesn’t need to be designed for “the average person’sunderstanding.” Of course, this doesn’t mean we throw human-centered design out the window; in fact, human factors as a discipline has its roots in aviation. But it is a paradigm shift from the likes of big tech companies who are focused on ensuring that anyone can use their newest phone, or latest electric scooter. We can design our systems around a specialized user, who has expectations about how that system should behave.

One of the more interesting aspects of being a human factors engineer in aerospace is that we are generally the only discipline that is concerned about the pilots and operators themselves. Avionics, NAV, subsystems, propulsion? All concerned with hardware and software on the jet and its performance relative to a spec. And all of that is absolutely required to make a great aircraft. But one thing I’ve learned over the years is it doesn’t matter if the jet can pull 20Gs if the pilot passes out at 10. Understanding the human limitations of a system is crucial to maximizing the potential and not over-engineering the platform beyond its usefulness and practical application.

This is the perspective that I have tried to bring with me to Celedon and Davinci. Most traditional MBSE tools are quite powerful. They can do many things, helping to keep track of programs, design elements, requirements and so much more. Once you have it set up, they can be truly beneficial to any project you could imagine. But there lies the problem. Getting it set up is a pain. For teams who have never worked with MBSE before, the learning curve is intense. On top of that, the actual transition to the program is a nightmare, needing dozens, hundreds or even thousands of hours to manually input all the system data into the program. Tracking down truth sources in documents and transcribing them into a new application is often an entire team’s full-time job. Once the model is fully set up (if it ever actually is), maintenance of that model is never ending and time consuming. And many of these pain points are true even for experienced MBSE users. There is simply no way around it; humans are the limiting factor in the current MBSE software paradigm.

As I’m sure you read in Jared’s blog post, we’ve experienced many of these problems firsthand, and that is what led us to creating Davinci. Engineering should serve the needs of humans, not the other way around. Imagine a system where you can upload the truth source and have it parsed into model objects for you automatically. Where you can ask questions in plain English and get answers directly from model data. Where deliverables and model documentation can be automatically produced on the fly by a digital agent with expertise in your system. This is the goal of Davinci, and so much more. 

You shouldn’t need specialized training or certifications to make digital models. Our ideal user is not someone specialized in model creation who wants to spend hours upon hours learning clunky interfaces and tools with the hopes of someday reaping the benefits of digital design. Our ideal user is you; someone with novel ideas; creators; designers; engineers; someone who will shape the future. Someone who, through a few simple sentences, can create a detailed, fully integrated and validated system using Davinci. Because you already know what you want to make, now let Davinci help you make it.

---