Computational Engineering: how a new paradigm will soon change everything
--
Today, engineering design and testing mistakes are now causing aeroplane failures and repeatedly costing lives, and emissions scandals are just the start of our issues. Modern-day engineering has become is slow, expensive and painful. What used to feel like invention, now feels like a grind. Our designs are now so complex, that industry needs to spend years and millions to deliver incremental updates.
At the same time, there’s so much to be excited about. We’re currently going through a generational move from Internal Combustion Engines to Sustainable Electic Vehicles, and are now pushing to becoming a multi-planetary species.
Combine these needs, the hard push for new generations of products and the rise of 3D printing, the maker community, cloud computing and mass customisation, and you can tell that we’re around the corner from a big change in the industry.
Everyone knows that the industry is set for something big, and computational engineering design will be the spark that sets it all off.
Computational Engineering (CE) is a new approach to engineering design that massively increases the scale and complexity of designs that engineers are able to invent. CE gives a single engineer the power to operate with the resource of a whole team and gives teams the ability to invent things that we’ve never seen before. CE will change the landscape of the industry, in ways we can't begin to imagine, but it’s going to be very different from what anyone is currently imagining.
We’re entering a new era for engineering:
Some quick history: From the 1960s to the 2010s, the industry went through a change of how we did engineering. As projects began to get more and more complex (think the early Ford Mustang to a modern-day Bugatti Veyron), our old approach no longer scaled. We needed to find a more scalable way.
To solve this, we moved from Manual Engineering (doing work on pen and paper) to Digital Engineering (doing work on a computer). Same work, moved to software. We went from blackboards to spreadsheets, drafting to CAD and file cupboards to SharePoint.
With each one of these changes, we found the joys of digitisation. We could make designs parametric, could instantly share and collaborate on ideas and replicate designs at zero cost. This move enabled the modern world as we know it. Modern-day products, whether it be the SpaceX Falcon 9, or the Airbus A380, are so complex, that they would be near impossible to design using pen and paper.
Over the last 50 years, complexity has once again increased massively. And it's not just complexity in a single domain, but lots of different domains (Structures, Electrical, Aerodynamic, Chemical, and Regulation Changes) are all deeply intertwined. This makes our current process extremely hard and painful. It’s now so expensive and painful to design, we’re forcing engineers to work overtime just to keep up with the design revisions and iterations. Once again, the tools that we use to design products are barely keeping up, leaving the burden on engineers to work harder, instead of smarter.
Right now, the industry is going through another transition. It’s a more subtle transition, but the impact will be hundreds of times larger. It will not just change what we can design, but will fundamentally change how we design.
The transition is from Digital Engineering to Computational Engineering.
So what is Computational Engineering?
Computational Engineering means that rather than doing the work, the computer does the work for you.
This means that the Computer works as an ally to the Engineer rather than software that you need to fight with. The Computer does a lot (if not all) of the heavy lifting, such as the implementation, calculations, generating Geometry, and validating the Design. The Computer does this in the background, letting the Engineer focus on what Engineers do best: creating new Ideas and exploring the Design.”
In effect, there’s a little army of incredibly fast and powerful workers inside the computer, doing your bidding. Unlike generative design or solvers, engineers are in control of the process — the inventors.
The transition from Digital Engineering to Computational Engineering will affect all three pillars of engineering: Design, Validation and Manufacture.
We will move from manually drawing pictures of our ideas in CAD, to defining what we want: the computer will do all of the manual execution work in generating the CAD for us.
We will move from setting up and running simulations manually to telling the computer what we want to verify: the computer will set up, run, interpret the simulation results, and suggest changes to improve the design.
We will move from manually programming CNC toolpaths, creating set-up sheets and work instructions, to automated machining and 3D printing.
With all of the manual execution work done by the computer, engineers will get to focus on the engineering work that matters— the invention.
Same story, other industries
Although this shift sounds scary, it’s very natural. This transition has happened in almost every other field of creative design and enabled much of the modern world (Software Development, Gaming, Animation, VFX, and IC Chip Design).
It’s always the same story — we do stuff manually on pen and paper, which works for some time, but then complexity rises and it becomes too difficult to continue. We then move to the digitisation phase, where we take the old, trusted process and bring it to software, making the process smoother and more efficient. This second phase means that we can dare to make even more complex products until it, again, becomes too complex and hard, until we move into the third phase: the Computational Era.
Programing languages, modern video game engines, animation engines and chip design are all phase three industries. Engineering design today is moving from phase two to phase three.
This is a fundamental change in how we do things and its hard to imagine how much of an impact it will have on the industry. Just like it’s hard to explain to game designers from the Pong and Mario era, the impact that games engines would have in bringing us to the modern call of Duty era.
The shift into Phase 3 for Engineering will enable cities on Mars, massively more complex Robotics, and will unlock the Starship Enterprise.
But have no doubt: this is just the start.
What happens to Engineers?
A common mistake is to think that if the computers are now doing the work, that the engineers would for some reason become less valuable, or go away completely — it is, in fact, the opposite.
Engineers do not get automated away — engineers get superpowers. The secret is that the role of the engineer changes, from the executor (1st person mode) to the architect (3rd person — with the computer becoming the executor). The way we typically do this is by programming, but programming doesn’t need to be scary, ugly, or even code-based. Like any transition, we need to learn new tools, but those who do early will set the pace for what is to come.
Another implication of abstractions are that they massively lower barriers to entry. Today a 10-year-old kid can create a new python script, write 100 easy lines of code, pull libraries from Github, and design an app, a game or a neural net in hours, with the computer generate millions of 1’s and 0’s for them in an instant. What would have taken a team of world-leading punch card operators a year to do in the 1950s, is now a kid and an hour.
For inventors, this gives them the ability to create new products and compete with the major players with a fraction of their resource. Every abstraction to date has enabled a new wave of inventors, technologies, products and companies and it will be no different here.
Why now?
Surprisingly, CE has been operating quietly at the fringes for quite some time— people all over the industry have been using their own little hacks for the best part of 20 years. We’ve all heard of the people in our organisations that that scripted his way out of doing big chunks of repetitive work. Or the Engineer that builds Python loops to design product variants or automate simulation, letting the computer do their day job overnight. CE in some shape or form has been happening at the fringes for two decades, ever since CAD and simulation packages opened up their API’s.
So if little pockets of engineering are converting, so why is it important for the bulk of the industry to move over now?
Today, design and testing mistakes are now causing aeroplane failures and repeatedly costing lives, and emissions scandals are just the start of our issues. Companies are now facing new economic realities with rock bottom oil and gas prices.
I’m lucky enough to speak to lots of different industries, and I hear the same stories over and over, engineering costs are too high, we’re behind schedule, and companies need new processes to simply keep up. Its hard to make ends meet with the current approach.
At the same time, there’s so much to be excited about. We’re going into a new era for aerospace and automotive. We’re currently going through a generational move from Internal Combustion Engines to Sustainable Electic Vehicles, and are now pushing to becoming a multi-planetary species.
Combine these needs, the hard push for new generations of products and the rise of 3D printing, the maker community, cloud computing and mass customisation, and it’s clear that we’re around the corner from a big change in the industry.
We must find ways not just to keep up, but to keep pushing the frontier in these uncertain times. On the fringes, we’ve already found a way to work unbelievably efficiently, leveraging computers to help engineers do more.
This approach, although is thousands of times more powerful than traditional methods, is incredibly inaccessible. It requires us learning hard software engineering languages, programming using terrible CAD API’s, and using inconsistent macros. Although modern startups and cutting edge companies have in house automation efforts, this approach is too hard and too painful for everyday engineers.
It’s more important than ever that now, we take this approach, and find a way to make it accessible to every engineer in the world. This is a tooling problem.
If we’re able to build the right tools, and make them easy and accessible, we have a chance to permanently change the course of our industry, and perhaps even, the human race.
The rise of the open-source design:
There’s one final piece to every successful transition — the open-source community.
The open-source community helps proliferate innovation, invention and education across an industry. Without open-source, we would not have Linux, Android, most cloud technologies, and hundreds of thousands of games. Without video tutorials and content, there would be a fraction of the no of self-taught engineers. Most new projects today (both hobbyist and enterprise-grade) are built using open-source technologies and Engineering Design will be no different.
The industry is always responsible for products and services, but often the open-source community invents, shares and standardises the core infrastructure we all build on.
The Engineering Company
We at The Engineering Company, know how important it is for this transition to happen now. TEC was founded to accelerate the industry to transition to CE — what typically takes a generation, we believe can happen in a matter of years. We’re building the Computational Engineering design paradigm, the first easy and intuitive design tool for engineers, the underlying technologies that enable it, the first sets of libraries and tutorials to get us started, and of course, the laying the seeds for the open-source community to thrive.
If you share our vision of the future and want to help us lay the foundation of the open-source community, message me.
Disclaimer; This is a short introduction to computational engineering. I have been making generalisations and oversimplifying in this post to communicate the core idea. I’ll be going into more detail in future posts.
— — — — — — — — — — — — — — — — — — — — — — — — — — — — — —
