Opinion
In an age of AI, climate adaptation and accelerating change, engineering education must return to first principles.
People often imagine engineering as a world of equations, calculations and technical systems.
Engineers are often portrayed as people fascinated by mathematics, machines, materials, and technology, yet somehow removed from history, society, ethics, and the humanities.
As a structural engineer, I understand where that perception comes from. Our language is technical. Our work is expressed through drawings, specifications, codes, calculations and digital models. A simulation can produce results that look beautifully precise.
Yet every engineering decision has always involved people.
How safe is safe enough? How should scarce resources be used? How should systems behave when conditions exceed what was expected? How do we protect communities, environments and public life over the long term?
Engineering has always been socio-technical. We have simply not always explained it that way.
Engineers Australia’s 2025 exposure draft of the National Competency Standard for Engineering makes this more visible. It identifies ten practice domains and six engineering mindsets across the profession and gives increased emphasis to socio-technical knowledge and skills, ethical decision making, communication, collaboration, management, lifelong learning, systems thinking, sustainability, risk, diversity and inclusion, and human centred practice.
This does not mean engineering has suddenly become human centred. It means the profession is now making explicit dimensions of engineering practice that were always present but often treated as implicit.
That creates an important opportunity for engineering education.
The challenge is no longer to argue that these capabilities matter. That question has largely been settled by the profession itself. The harder question is how to design engineering curricula so that judgement, responsibility and socio-technical integration are developed, practised and assessed across the student journey.
Technical knowledge still matters deeply. Engineering graduates need mathematics, science, modelling, analysis, design and disciplinary depth. A civil engineering graduate, for example, still needs to understand structures, materials, geotechnics, water, transport and infrastructure systems.
But technical knowledge alone is no longer enough.
Today’s graduates will enter a profession shaped by artificial intelligence, climate adaptation, ageing infrastructure, resource constraints and changing public expectations. They will need to use digital tools critically, work across disciplines, manage risk, understand lifecycle consequences and make responsible decisions where there may be no perfect answer.
The curriculum challenge, therefore, is not simply to add more content.
It is to ask a deeper question: how do we educate engineers who can exercise disciplined judgement under uncertainty?
Good engineering is not blind certainty. It is disciplined judgement under uncertainty.
Every model depends on assumptions: loads, material properties, supports, boundary conditions, construction quality, deterioration, patterns of use and events that may never occur in a designer’s lifetime. Every design decision involves choices about safety, risk, resources, context and consequences.
That kind of judgement does not emerge automatically because students have completed enough units. It must be intentionally developed, practised, assessed and reinforced across a course.
This is where many curricula struggle. Knowledge is often organised into discrete subjects, while integration is deferred to a final year capstone or assumed to occur naturally through good teaching. In sequential or block delivery models, the challenge can be sharper because connections across units cannot rely on timetable proximity. If we want students to use prior knowledge later, we must design deliberate points at which that knowledge is retrieved, applied, and defended.
Integration cannot be left to chance. It must become a structural property of the curriculum.
In our recent civil engineering curriculum modernisation work, this principle has been central. The aim has not been simply to add professional skills, sustainability language or digital tools. The aim has been to design a whole of course architecture in which technical learning, professional judgement, industry engagement, AI aware assessment, regenerative thinking and project capability are connected across the student journey.
One way to achieve this is to treat project-based integration not as an additional assessment task, but as an organising spine across the course. Carefully designed integrated projects require students to apply knowledge, state assumptions, use tools responsibly, consider stakeholders, manage risk, communicate trade-offs and justify decisions in context.
The point is not simply that students “do projects”. Many engineering programs already include project work. The deeper question is whether project work is deliberately connected to the technical curriculum, assessment design, professional standards and graduate capability development. When designed well, integrated projects become the place where technical depth and professional judgement meet.
This also changes assessment.
If we want students to develop judgement, we cannot assess only final answers. We need to see the reasoning that produced them. Students should be asked to document assumptions, compare options, verify digital outputs, explain limitations, reflect on risk, respond to feedback and defend decisions orally and in writing.
Design logs, staged milestones, brief checkpoints and peer review can help make student reasoning, authorship and professional decision making visible.
This is especially important in the age of generative AI.
AI should not be treated as an external disruption to be managed at the margins. It is already part of the professional world into which students will graduate. The question is not whether students will use AI and digital tools. They will.
The educational question is whether they can use them critically and ethically, with engineering judgement.
That means students need to learn how to verify outputs, compare AI supported suggestions against first principles, identify assumptions, recognise limitations and explain where human judgement remains essential.
If we teach only tools, we prepare students for obsolescence.
If we teach first principles, judgement and responsibility, we prepare them for uncertainty.
This is where engineering education must hold two things together.
On one hand, we must prepare students for technological change. Software will change. AI will change. Materials, methods and professional expectations will continue to evolve.
On the other hand, we must ground students in principles that endure.
Gravity has not changed. Equilibrium has not changed. The need for safety, integrity, responsibility and public trust has not changed.
First principles matter because they give students a way to think when tools change, when data are incomplete, when assumptions are uncertain, and when competing obligations must be reconciled.
Structural engineering offers a useful example.
One of the most humane principles embedded in structural design is ductility. To someone outside engineering, ductility may sound like a material property. To a structural engineer, it carries a deeper lesson.
A ductile structure does not pretend extreme events will never happen. It is designed so that, when stress exceeds ordinary expectations, failure is not sudden and catastrophic. Instead, it can deform, absorb energy and redistribute load, giving visible warning before collapse becomes possible. In practical terms, that warning matters because it can give occupants time to recognise danger, evacuate and survive.
Resilience, in this sense, is not the absence of deformation. It is the capacity to respond under stress without catastrophic failure.
There is a curriculum lesson in that.
We should not design engineering education for a stable world that no longer exists. We should design it for graduates who will encounter uncertainty, complexity and competing obligations. They will need technical knowledge, but also judgement, ethics, communication, systems thinking, risk awareness, cultural responsiveness and responsibility for long-term consequences.
These capabilities should not be treated as soft additions to a technical core. They are part of responsible engineering practice.
That requires a shift in how we think about curriculum.
We need whole of course architecture before individual unit design, because students experience a course as a learning journey, not as a set of isolated subjects. We need to make integration assessable, so professional judgement appears in rubrics, milestones, project artefacts and feedback, not only in graduate attribute statements. And we need collaborative curriculum governance, because integrated learning cannot be delivered through isolated unit ownership.
The purpose of engineering education is not to train students to reproduce yesterday’s solutions using today’s software.
It is to prepare them to make responsible decisions in a world none of us can fully predict.
Engineering has always been socio-technical. Now we have the language, standards and urgency to teach it intentionally.
The future does not need graduates who can only use tools.
It needs engineers who can judge when tools are useful, when they are limited, when assumptions matter, when people are affected and when responsibility cannot be outsourced.
That is the deeper work of engineering education now.
Professor Zora Vrcelj is Head of Built Environment and Engineering at Victoria University