Skip to the main content.

2 min read

The Role of AI in Modern Systems Engineering for Safety-Critical Processes

The Role of AI in Modern Systems Engineering for Safety-Critical Processes
The Role of AI in Modern Systems Engineering for Safety-Critical Processes
4:21

Visit our YouTube Channel for more content here.

The Role of AI in Modern Systems Engineering for Safety-Critical Processes

Using AI to Enhance Safety-Critical Systems Engineering: Insights, and Strategies

Hi again, this is Steve Neemeh creating a series of videos on creating a high-performing low-cost engineering organization for the 21st century.

In past videos, we talked about measuring performance at every step, the impact of data analytics, and also how to use AI to make your project managers and project engineers effective.

Today's topic is systems engineering and systems engineers in general, specifically targeted towards safety-critical applications or ones that have some regulation.

It can be argued that systems engineers are some of the most important people in the organization. They architect, they define, and then they bring together the hardware and software into a final solution or technical solution in a safety-critical space, whether it be aerospace with ARP 4754 standards or automotive with the functional safety ISO 26262.

It’s even more critical because they have a specific role with specific deliverables.

In my past life, I was challenged by having so many systems engineers that our costs of engineering were just escalating beyond the point of profitability. The observation that was made was most of those systems engineers were managing processes and delivering documents, while a very small number of those were architecting and creating the next-generation product.

Now, going back to the theme of a previous video of getting engineers to innovate, I'm going to list out a series of documents, deliverables, and objectives that I believe can be automated out of the process altogether.

Now let's break that down into the left side of the V and the right side of the V. I still believe, at this point in our transition towards AI, that we need people to innovate and we want people to innovate. The right side of the V, however, can be completely automated.

So in that vein, here's a list of deliverables, both from ARP 4754 as well as ISO 26262, which are required for the delivery of a product in safety-critical applications that today we hire four-year-degreed engineers to do.

Let’s start with requirements review and requirements consistency checks. You've got a requirements document, you're forced to develop requirements standards, you're required to develop a review process and a review deliverable, and we hire engineers to do that. That’s a language model problem and a language model solution.

Interface control documents. You've got a schematic, you've got software. You can draw out the interfaces, the signals, signal names, and all the information that goes with it and populate an interface control document. I have spent time doing that kind of work in a past life.

Also, test artifacts and test cases. You've got a requirements document. You can auto-generate test cases with boundary conditions and a document that goes with it and have it automatically update every time a requirements document gets updated. Right now, we hire engineers to do that and we offshore the review of that.

Your safety analysis results. You've created an FMEA, and you've created a fault tree. If an engineer is spending an hour formatting that in Word, you're wasting that hour. And if you've listened to some of the previous videos, hopefully, you've armed them with an AI tool that helps them use the FMEAs and fault trees from previous applications.

For systems engineering in general, there's an entire effort called Model-Based Systems Engineering and graphically documenting using UML and SysML the actual requirements and potentially auto-generating those requirements.

And finally, just general knowledge reuse. Systems engineers are required to architect the next-generation systems. Hopefully, they are armed with knowledge, know-how, and features of the previous 50 systems that were created in your organization.

That's just a small snippet of what can be automated in today's world.

If you like these kinds of videos, go ahead and subscribe and like, and maybe we can create a forum where we share these ideas and exchange information on the topic.

That's your $1,000,000 minute. Thanks.

Why Traditional Project Management is Broken and How AI Fixes It

Why Traditional Project Management is Broken and How AI Fixes It

{% video_player "embed_player" overrideable=False, type='hsvideo2', hide_playlist=True, viral_sharing=False, embed_button=False, autoplay=False,...

Read More
The Entwined Futures of Artificial Intelligence and Autonomous Vehicles

The Entwined Futures of Artificial Intelligence and Autonomous Vehicles

The Entwined Futures of Artificial Intelligence and Autonomous Vehicles In every era, automotive manufacturing has always had some degree of...

Read More