So more and more tools are being built for running automated simulation workflows, to power design of experiments and optimisation algorithms, and to generate enough data to enable AI. To make use of this data, engineering teams also need powerful platforms for understanding it and which can work within a collaborative, iterative design process.

But to get this data in the first place, you need to automate a full pipeline from generating geometry to post-processing simulation results.

The most visible steps get a lot of attention: generate and simulate: creating a valid geometry, and then spending significant compute time on running physics simulations. Either of these steps breaking or slowing down is a clear target to improve.

But we’ve been doing this with real engineering teams across industries for years (see our case studies in bioreactors, robotics, or architecture), and the bottleneck to making use of this pipeline isn’t either of these steps. It’s the cognitive challenge of building parametric versions of all the other steps that an engineer usually does manually.

The cost of setting up

In all of the engineering problems we’ve worked on, setting up the pipeline means: writing automation logic, connecting tools together, debugging, and thinking through how parametric changes affect everything from geometry to post-processing of simulation.

All of these steps need to happen regardless of what geometry tool you use, which scripting language, or what simulation package. More robust geometry kernels (like implicit models) don’t produce a broken BREP when you push parameters to unexpected places, but that doesn’t reduce the core cognitive challenge: deciding what should vary, what stays coupled, and where the boundaries are. Additionally, if you want to use physics AI, you still need all of this first – plus model training and validation. During inference, you can often skip meshing and other preparation steps, but still need this full pipeline for training.

While you’re going through all of these steps, there is a key question to answer: “by the time this pipeline is set up, what kind of design iterations will we be exploring”? You’re trying to predict the future, while the design is actively changing.

Only the changes compatible with the automation you’ve set up can be explored. But because of the iterative nature of engineering design, by the time the pipeline works, the design has already moved on. The more iterative you are, the better your engineering process; but the faster the design evolves the more challenging this prediction becomes.

This is why most tools for automating this pipeline create monolithic workflows that only the author can iterate on. The computational engineer becomes the single point of dependency.

One emerging response is to try and build AI agents that handle simulation setup on the fly. Agents can be incredibly powerful with the right tools and skills, but using them on every new design shifts where the expertise is needed rather than reducing it. Without a verifiable environment of known-good cases to validate against (or until we can completely trust AI’s output in engineering), someone still needs the expertise to judge when the agent’s choices are subtly wrong. And running a probabilistic agent within the pipeline itself isn’t an option when you need fully traceable results.

The challenge is to make agents work to speed up this automation pipeline in a way that supports the reality of engineering design iterations.

The parametric challenge

Setting up a parametric pipeline is challenging in itself, but engineering iteration rarely stays parametric. Real design changes aren’t “adjust this fillet from 2mm to 4mm.” They’re “split this into two components,” “try a different mounting approach,” or “the supplier says this material isn’t available.” These are new designs, not parameter tweaks.

These beyond-parametric changes aren’t edge cases — they’re the norm. If you’re working with an industrial designer, you might receive a render or just a surface, not full CAD. A supplier sends a scanned part in a different format. A design review produces an annotation sketched directly onto the model. They don’t fit parametric models, but they still need to be evaluated. In our experience, the computational engineer responsible for the parametric workflow often doesn’t have the time to adapt the automated workflows to fit these new directions.

For automated simulation pipelines to remain useful across actual design cycles, they have to accept design ideas that weren’t anticipated when it was created.

One approach would be to make the automated design space large enough to cover everything. This is what physics foundation models, or large geometry models, are attempting. But simulating physics isn’t about learning high-level concepts from more data; it’s about precise mathematical relationships in fundamentally chaotic systems, and LLMs haven’t yet demonstrated reliable geometric reasoning. Whether foundational models for geometry or simulation can be built in practice remains an open question.

Rather than trying to make the scope of automation larger, we reduce the setup cost so automation can evolve with the design.

Trust experts in simulation, accelerate geometry

Over many years we have made an observation that helps: simulation templates are far more reusable than parametric geometry models.

The criteria you evaluate against — fatigue life, static stress, mass, cost — stay stable across design iterations, even as the geometry changes radically. A simulation setup defines boundary conditions, meshing strategies, convergence criteria, and de-featuring rules. These define the problem, not the design. A single simulation setup can handle far more geometry variation than a monolithic geometry model.

In our platform, the simulation template remains human-defined because that’s where trust and expertise matter. We use AI to accelerate the expert creating these templates, but keep the expert accountable, and instead we speed up the rest of the pipeline.

Automation that evolves iteratively

To reduce the setup time for automation (in order to enable the data-driven decisions that engineering teams need to move faster), we focus on all the steps around generation and simulation. We’ve developed AI agents to write the automation logic itself for parametrising geometry, feature-detection, geometry-preparation, meshing, configuring simulation, and post-processing. This removes the need for an agent to reason about each design at runtime, keeping all automation logic deterministic and traceable.

We create a continually growing set of verification cases for the agents to work with from successfully simulated examples and all human-defined designs (specific to your problem, keeping all IP yours). This allows the automation to improve over time, regardless of how the design evolves, escaping the parametric bottleneck.

If automation fails on a novel geometry, the engineer does what they would have done anyway: uploads the geometry, assigns boundary conditions and mesh configurations themselves, and continues. The setup cost of this approach converges toward zero. You start by exploring and simulating normally — every correction you make along the way becomes a test case. By the time you come back to the next round of designs, the AI has used those cases offline to improve the automation logic. You don’t need to predict the future, because the system learns in real-time as you work.

One expert-defined simulation setup supports an entire design team exploring freely across novel geometries. Analysts review exceptions, not every run.

We’re building for teams where speed to market matters more than optimisation performance. Startups and scale-ups iterating weekly on fundamentally different designs, not refining the same one for months, including robotics, bioreactors, micro-mobility, fusion energy, wind turbines, the next generation of energy technology, and an expanding set of industries.

If these bottlenecks feel familiar, or if setting up generative studies hasn’t been practical yet, we’d like to hear from you.

This paradigm of manual iteration does eventually work. It’s trusted in engineering industries where bad design decisions can cause death, where strict regulations need to be met before any product can be used, and where expensive machinery and supply chains are needed for manufacturing. 

Yet, to meet humanity’s most pressing problems, our world needs re-engineering with novel solutions made from novel materials that push the limits of past experience. This creates an immense space of design possibilities for engineers to navigate, but also creates a huge opportunity for engineering. 

A new paradigm of iterating on designs is needed: one where designs are generated rather than drafted, where an engineer's intuition is superpowered by exploring many more possibilities, and where the whole team is able to make efficient and collaborative decisions in a fraction of the time.

From drafted to generated design

Manual drafting involves humans directly manipulating designs in CAD tools like SolidWorks, Siemens NX, or AutoCAD. Each design must be created by hand, one at a time. When designs are generated instead, engineers control high-level parameters and let the design details (geometry and otherwise) be created automatically.

You might have heard about or experimented with methods like:

• Topology optimization (what the term “generative design” in engineering often refers to).

• Computational engineering, through visual programming or writing code.

• Optimisation and exploration algorithms (like those in process integration and design optimization tools)

• Text-to-CAD AI

Methods like this let engineers work where they are strongest: handling complex trade-offs and inventing solutions, rather than repetitive tasks managing low-level details. 

While these methods show incredible promise, they haven’t realised their potential due to the lack of supporting infrastructure for seamless integration into organisations. As modern engineering systems grow more complex, real-world engineering remains unpredictable — and deeply human. The challenge lies in bridging the gap between sophisticated methods and the realities of messy, human-driven processes.

Engineering is increasingly complex

With new technologies, such as additive manufacturing, complexity is only increasing: compare a simple bracket designed for sheet metal to one designed for additive manufacturing from novel materials. 

Now imagine the increase in complexity when these new technologies are introduced to systems like cars, buildings, wind farms, infrastructure and everything else the future demands. This complexity makes manual iteration responsible for weeks, months or even years of delays: manually changing all aspects of a highly complex system is laborious, prone to error, and makes it hard to predict the knock-on effect of propagating changes.

Some tools of today see the pain of this complexity and try to manage it: requirements management tools, product data management (PDM) tools, bill-of-material (BOM) spreadsheets etc. Yet they are built for a single-design world and assume that a human should work at the level of complexity of the engineering system directly. 

Irrespective of how good they are (ask around and you’ll see that they are not engineers’ favourite tools), this central assumption inherently calls for more data management as physical products get more complex, rather than less. 

To keep engineers doing engineering, the assumption that a human works at the level of complexity of the system needs to change, and the tools built for this assumption need to be replaced.

Engineering is Human

Engineering complex products requires a team of people with a wide variety of expertise: product managers, business stakeholders, mechanical engineers, electrical engineers, aerodynamics engineers, simulation teams, clients, suppliers, part vendors, and more. All of these need to come together to align on the design of high-performing, low-cost, manufacturable parts and systems.

This means collaborative reviewing of design possibilities is central to the engineering design process. Today, engineers present a small number of manually created design possibilities to get reviews from many stakeholders with many conflicting goals. This then requires hard trade-offs to be made collectively, without enough data to realistically know the right direction. So engineers rely on intuition gained from experience, and the most experienced (or loudest) voice in the room usually wins. Then the original engineer goes away to manually redesign their component based on this feedback.

When a small variation on an existing product is being designed, this intuition might be good enough. When designing novel engineering products, making decisions without enough data leads to years of delays. 

Quantity as the driver of quality

Most generative software tools let engineers create a so-called “optimal” design that meets their high-level targets. This leads to comparisons such as “the generated design ended up 12% cheaper and 20% lighter than the manually created one”. Such comparisons with traditional methods, while already beneficial, miss out on some key advantages:

• You can generate many design options at once - for a similar cost & time to creating one.

• You can re-generate when the high-level requirements change, without manual iteration. 

With these, considering trade-offs becomes instantaneous. Today engineers manually explore different options, analyse each one, and present trade-offs in design reviews. Tomorrow the design review is where all trade-offs can be made in real-time — as easy as generating a single design.

Anyone can generate new options by simply updating high level parameters, making it easier for teams to respond to changing needs and requirements. The engineers avoid having to work directly with the low-level & complex detail.

The set of generated designs can then be used as the basis for the next iteration, giving a paradigm of iteration based on repeated generation of many designs rather than repeated manual changes of one design.

Imagine every question asked by a product manager, engineering expert, or customer being instantly answered by a clear impact: “improving the performance by 11% would increase manufacturing costs by 8%, shall we continue?”. How much faster could engineering systems be brought to market?

The potential of this paradigm to improve how engineering products are designed is large, but just having the ability for one engineer to generate is not enough. The rest of the process of engineering design has to improve with it for this potential to be realised.

Empowering engineering organisations today

It’s a journey driven by a thriving community. Companies like nTop, Synera, Simscale, Leap71, Hyperganic, ToffeeX, Zoo, and Grasshopper are empowering innovative engineers to generate and test designs in remarkable new ways. 

But these innovators often face scepticism from colleagues and suppliers, who find it hard to manage and trust the many generated design options. On top of that, integrating generative work into their organisation’s data and project management processes can quickly turn into a nightmare.

It’s time to change that—we're out to support innovative engineers that want to transform existing manual engineering into generative engineering, for the whole team, one step at a time, with engineering software fit for challenges of the next century.

We are calling out to all of these innovators: Generative Engineering’s technology is now available for early access. The engineering industry needs your input to test, criticise, and enable what we perceive as the future of engineering design. We do not believe in replacing engineers with software, we believe new software will enable engineers to use their expertise to start building tomorrow's products, today.

For early access to what we’re building…

Visit our website

Or reach out at helloworld@generative.vision.

Our human-manufactured world was designed and engineered during a time of abundantly cheap energy and materials. It gave us productivity, comfort, and confidence. Cars, for example, reinvented personal transport, but in time they also created shopping malls and eventually reshaped cities. Fossil fuels provided easy energy for heating and transport, but also feedstock for manufacturing plastics, textiles, adhesives, and chemicals.

But the world we designed for has already gone

When you sit in a meeting, travel across the country, or go to the shops, you’re part of an engineered world which now belongs to the past. Products, tools, transport, infrastructure, systems, energy, housing stock, jobs, cities, and even society — were designed for an era which has now vanished.

We no longer have abundantly cheap energy or materials. Climate change swept that world away.

We have spent a century emitting carbon into the atmosphere. The secondary effects and risks came into focus only gradually. We began to see that the externalities in the systems we’d created were in direct conflict with a sustainable future. They were beginning to unmake the planet too: the stable, persistent world we understood and relied upon has given way to one of change.

In short, the human layers we have built are no longer fit for purpose.

Can our world be radically re-engineered?

We live in an era of accelerating change, instability, and unpredictability. Therefore it must also be the era of us finally responding boldly: the invention of new solutions and technologies must be much faster, more flexible, and highly responsive.

We see hope. The first generation to grow up with connected devices has now come of age, and the result is a complete revolution in communication. Global population has doubled in the past 50 years, partly enabling this revolution with an abundance of eager adopters of new technology. Can this be mirrored elsewhere?

Opportunities everywhere

Every sector faces significant challenges. Meeting them will need energy plus ingenuity.

Energy and electricity

Electricity generation from fossil fuels is becoming less cost-competitive as the capital costs of renewables drop. However, the distributed and intermittent nature of renewable energy generation means distributed storage is necessary. We will need utility-scale storage, and to keep energy locally in domestic batteries and electric vehicles. Today’s electricity grids need to be more flexible and resilient. Do we need microgrids, better energy supply and more intelligent demand management? Or completely new forms of renewable energy?

Land use and food production

Water supplies are under significant pressure. Extreme weather, disease, and pests lead to crop losses. Yet fuel crops, meat production, carbon credit forestry, solar, and wind farms all increase competition for land. In the background is biodiversity loss, at a cost we won’t even understand until it’s too late.

Growing populations generally require increased food production. But land degradation, nutrient leaching, desertification, and salination are reducing the amount of available land. Fertilisers have filled the productivity gap in recent years, but they are mainly derived from fossil fuels and non-renewable resources. Regenerative agriculture must scale and thrive.

Housing, buildings, and heat

Heat pumps are starting to make inroads into fossil-fueled home heating but in the UK alone, 20m+ homes are under-insulated. The low building replacement rate in the UK and other countries means that most of these buildings will still be in use in 20 or more years. Alongside a critical need for lower carbon energy generation, these buildings need a massive programme of insulation retrofit.

Transport

The gradual electrification of everything is driving up demand: electric vehicles have grown market share in many countries, but EV charging infrastructure often lags. A bigger question is whether cars are the right tool for the job. The average car or van in England is parked for 95% of the time, and when they are driven they’re used for single-passenger journeys 65% of the time. There is an explosion of smaller “micromobility” vehicles including eBikes, electric scooters, pods, and neighbourhood EVs. They’re growing at a faster rate than cars because they are better suited to short and often urban journeys.

But, we are now starting to see negativity thrown at electric transport, mainly as vehicle manufacturers are unable to sell their EVs without making a huge loss per product. The bill of materials cost and legacy engineering currently prohibit this.

Beyond personal transport, electricity, and hydrogen will be key factors in heavier transport and shipping. But is this right? Why build a megafactory if the capital expense of this drives demand for shipping products to all corners of the globe in an unsustainable way?

A million more engineers

Today, using traditional creation, design, and engineering methods, we will need a huge number of makers, designers, scientists, engineers, and problem solvers to remake the engineered world for this new era that demands better.

And we need all of them to work with greatly increased productivity and impact.

Their challenge will be to work to crisis timelines because the need for action is so urgent. They’ll need to design systems in a much more integrated way because everything is now connected and interdependent.

Whilst building new systems, engineers will need to keep existing systems running until they can be transitioned smoothly and safely. We’ll need to change the engines while the plane is flying.

There are not enough engineers today to do this work — we will need millions more of them. But will traditional methods of training engineers close this gap?

UK engineering is ageing

Every year, the UK needs to add 124,000 engineers and technicians across the country, as well as another 79,000 roles that require a broad engineering skillset. The shortfall is 59,000 engineers every year, plus this gap will increase over time.

The Committee on Climate Change estimates that between 135,000 and 725,000 net new jobs could be created by 2030 in low-carbon sectors such as generating renewable energy, upgrading buildings and manufacturing electric vehicles. Other estimates put the skills gap in heat pump engineering at 24,000 by 2028, in offshore wind at 70,000 by 2030, and in nuclear at 173,000. The National Grid alone will need another 400,000 energy workers to achieve net zero.

However, engineering apprenticeships are disappearing, and engineering isn’t perceived as an industry with exciting career opportunities. Many left the workforce during the downsizing after the 2008 financial crisis and didn’t return. Moreover, in seeking opportunity, why work in hardware when software can be far more financially rewarding?

It gets worse though, because the average age of engineers in the UK is around 55, and they’ll retire quicker than we’re currently replacing them. This decade, 20% of the engineering workforce will retire.

The US needs 11 million engineers

There are similar workforce gaps in other developed world countries. Let’s take as an example the problem of electrifying the United States.

In 2020, Rewiring America estimated that electrifying America would require 25 million workers at its peak, up from the 1.8m current jobs in the US energy sector.

If we assume that half of these 23 million new jobs will need university-level training, the US would need 11.6 million additional STEM graduates. In 2019–20, the US education system only produced around 430,000 STEM graduates at Bachelor level and 172,000 at Masters and doctorate level. However, in 2019 only 28% of US STEM graduates were working in a STEM job, and the US already had several million STEM job vacancies that needed to be filled.

So in an ideal scenario in which every new STEM graduate was free to focus solely on electrification, it could take 40 years to train enough scientists and engineers to meet the need. But this need is only one part of the bigger picture: it would take hundreds of years to train the scientists and engineers needed to achieve climate transformation.

And we have more like 10–12 years to get the whole job finished.

A programme to electrify America with maximum effort would need the same intensity and focus as the Second World War or COVID mobilisations. Is this likely?

Ideas everywhere, massive impact

We need a faster and better way to create engineers. One that isn’t competing for talent with the existing system, but that is incremental and parallel to it. And we need to make all engineers more efficient and productive, regardless of their training. They will need vastly more powerful tools.

As we surpass 8 billion people on the planet, imagine the diversity of ideas that exist and are generated every day. Ideas are an infinite resource. Our collective intelligence is vast. However, to be applied at the speed required, it needs to be concentrated, shared, and reused in the most efficient way possible by the widest audience possible.

Imagine if we had tools which can augment and channel our collective intelligence to quickly and effectively engineer the new systems we need. Today’s standard engineering process is expensive, highly complex, and time-consuming — the underlying reason many hardware companies (especially hardware companies building sustainable products) are regularly failing.

For example, in its initial attempt to become sustainable, the automotive industry currently shoehorns batteries and EV tech into legacy ICE vehicle designs, leading to inefficient and costly products and bill of material costs that far exceed the price to a customer. Yet the capital expenditure and time of engineering a sustainable platform from scratch is too big a risk to automotive manufacturers. Leading to mass reduction of sustainable commitments throughout the industry. This must change.

No engineering graduate today wants to work on legacy and highly polluting products. Engineers like to question everything yet are unable to fully push the boundaries of innovation because of cost, and most of us have ideas we’d like to see created. The diversity of knowledge behind modernisation will push forward innovation.

What if the time from idea to manufacturable product becomes hours rather than years? What if the risk-averse behaviour of manufacturers became history due to highly efficient engineering?

We have begun to build the answer. The challenge Generative Engineering is tackling is big, we won’t make it alone, and we need a firm realisation from us all that change must happen.

Help us engineer a future that works

helloworld@generative.vision for a demo