The five themes
to shape your next core

Earlier, we identified the three main factors influencing decision-making for CxOs and engineering leaders in particular: the pace of change, the convergence of digital and physical innovations, and unpredictable external forces. These factors underscore the importance of having a thematic focus to guide our engineering and technological efforts and the way we prioritize.

At Capgemini Engineering, five key themes anchor our strategic direction; we believe these themes are universal. These themes emerged organically; across the different roadmaps our teams developed in all the disciplines where we work, we spotted commonalities, signposting us to crucial topics that are everywhere in our customers’ worlds and that influence them daily. These themes are interconnected, highlighting the multifaceted nature of modern engineering challenges. Together, they are the underlying forces driving the evolution of engineering and technology, and they impact every field in which we work.

1.
Organic
engineering

Organic engineering is inspired by the resilience, efficiency and autonomous behavior observed in natural systems. This theme focuses on designing products and systems that are both autonomous and efficient, much like nature itself. The aim is to create engineering solutions that can adapt and evolve, ensuring long-term sustainability and resilience in the face of disruptions, such as those seen during the COVID-19 pandemic. This approach is evident in innovations like additive manufacturing, where nature-inspired lattice structures enhance product strength and flexibility, or in the emergent behavior of multi-agent systems.

Bio-inspired engineering technologies: The process of creating ideas, methods and technologies that plan, construct and manage in a manner akin to that of nature is known as ‘Bio-Inspired Design’, and it has the potential to revolutionize both market possibilities and challenges. These technologies range from nature-inspired robotics to replicating nature-generated materials (e.g. spider silk) to biomimetic energy harvesting (e.g. photosynthesis).
Biomanufacturing: Biomanufacturing, or the design and engineering of biological systems to create and improve processes and products, offers new ways to produce almost everything that humans consume, from flavors and fabrics to foods and fuels ^[11]. This is because, in theory, microorganisms can make many of the things that industrial processes currently manufacture. Here, the availability of raw resources may no longer limit supply. Businesses could design and produce an endless number of items from the ground up, cell by cell.
Additive layer manufacturing: As hybrid products – products that combine additive manufacturing with conventional manufacturing – become more prevalent, a large portion of printed goods produced using additive manufacturing will be made of several materials and/or have integrated electronics. This opens up a far wider range of applications. This potential range of uses will be impacted by additive manufacturing's capacity to produce goods with sophisticated features (e.g. multi-material construction or embedded electronics), which also suggests a greater influence for other projects. Experts estimate that if this technology proves to be successful, it will have an even bigger influence on manufacturing than current ‘standard’ additive manufacturing, and that businesses with smart manufacturing abilities will have significant commercial potential.

Some examples of organic engineering can be:

Furthermore, our engineering experience has allowed us to identify several fundamental, broadly unified technological and scientific paradigms that are becoming increasingly important in enabling Organic Engineering work. These paradigms, outlined below, will therefore be essential for any engineering leader who sees this theme as relevant to their own company:

Artificial intelligence.
Self-assembly, 3D nanofabrication and improved 3D printing.
Synthetic biology (cell-free, engineered cells and synthetic cells) and organoid technologies.
Robotics and organismal biology, including collective behavior and swarms.
Novel materials design and manufacture.
Computational modeling and theory.

We have also identified a subset of convergent tasks or functionalities that are important for Organic Engineering (this is a non-exhaustive list):

Fast prototyping, which increases speed, scale and quality, whilst decreasing cost.
Interfacing across materials and facilitating modular materials compatibility.
Spanning scales from nano to micron, including in biological systems.
Developing materials and systems with specified/optimized functionality, e.g. regenerative properties, sensing, self-healing, and the ability to retain memories, compute and evolve.
Discovery and engineering of degradative and biocatalytic systems.
Prediction of protein function based on sequence, resulting in improved accuracy and speed.

Pressures on energy production and distribution have a higher profile ^[12]: regulatory pressure on the use of fossil fuels, coupled with increasing market volatility and the commoditization of alternative technologies drives several opportunities, of which engineering leaders must be mindful as they build company priorities:

A resurgence in nuclear power, focusing in the short term on the life extension of existing stations, together with modular reactors (AMR/SMR), is a concrete trend, although new product development is understandably slow. The ultimate vision presented by fusion power is attractive, but remains in its early stages of engineering production. To reach their full potential in the world's energy mix, both fission and fusion must overcome obstacles including public opinion, legal restrictions, and economic feasibility in the upcoming ten years.
Hydrogen as a means of energy transfer and use is perhaps less politically difficult, but faces many technical and economic challenges. In addition to direct use for energy production or industrial processes, transportation is increasingly anticipated to be a primary user of hydrogen through derivatives (e-ammonia, e-methanol, etc.). Creating a hydrogen value chain that is safe, effective and reasonably priced will require innovation and industrialization throughout the lifecycle.
Any energy change – electrification or hydrogenation – will require amendment to the processes that use it, and a corresponding business case. Resources formerly considered commodities are now crucial to the resiliency of businesses. Interactions between costs, risks and government interventions by regulation or subsidy will need to drive the renovation of legacy facilities and the creation of innovation alternatives.
Even at an operational level, changes will be required to allow energy to be managed and distributed: infrastructure that is already close to capacity will need to be extended and managed to cope with variable energy sources, new storage mechanisms and distributed generation. Smart management systems will be essential.

2.
Resource
revolution

Engineering leaders must accept that the fundamental subjects of engineering – materials and energy – are changing, under pressure from environmental concerns, strategic constraints and technological demands for ever higher performance. The resource revolution theme addresses both the energy transition and the increasing need to engineer materials to meet specific requirements that natural materials cannot fulfill.

Demand on physical materials and processes is less driven by external pressures but is motivated by opportunities and market needs: new technologies and processes allow for the creation of new materials, and thus new products and new markets. Examples include:

Nanomaterials, interpreted broadly, are any materials with units of size in the order of nanometers. Of particular interest are active nanostructures which change some aspect of their state (morphology, mechanical, electrical, magnetic, optical or biological qualities, etc.) in operation. Mechanisms for delivering active chemicals, while overcoming some chemical or biological barrier, in medical applications would be one class of examples. Others include electronic devices, molecular machines, light-activated molecular motors, devices utilizing nano-fluidics, and nano-electromechanical systems (NEMS).
Metamaterials are artificially created materials with extraordinary properties – acoustic, electrical, magnetic and optical – due to their structure. Arrays of resonators can be arranged in them to manipulate sound or electromagnetic waves in ways that are not often observed in nature. Due to their adaptable dielectric characteristics and adjustable responses, they offer remarkable versatility in various applications, including telecommunications, sensing, optics, and imaging.
Multifunctional materials which are designed to contribute to different functional needs. They may serve as energy storage systems, adapt according to their individual circumstances of operation, or increase sustainability by facilitating innovative assembly and disassembly procedures. Such applications are made possible by an understanding of function-structure links and the development of expertise in the system approach for multifunctionality. This must be supported by a multidisciplinary approach, including modern production techniques that leverage digitization, as well as computational engineering methods for virtual material design.

Together these axes place strong constraints on an organization’s ability to recognize value in fundamental research and innovation, construct and evaluate business cases (in the context of rapidly changing markets and complex regulatory environments) and integrate the results into their products, processes and operations. These are all things that fit within the purview of engineering leaders and, therefore, these leaders must monitor the following elements as part of their decision making:

Given the breadth of this theme, companies need a clear direction on where to focus their attention. A clear understanding of the impact of these revolutions on the Core of the company will help sharpen the efforts of research and technology teams.
Continuous monitoring of the results produced by research teams in the domains that impact our Core. The vast amount of literature and publications requires a structured approach to discovering what is happening out there.
Good understanding of physics and chemistry at the fundamental level, as these energy and materials revolutions are based on them. Given the broad spectrum of different disciplines involved (atomic and molecular physics, properties of materials at extreme temperature ranges, etc.), we recommend establishing partnerships with academic institutions that have the breadth and depth of knowledge needed to interpret the results of the ever-increasing research. Indeed, this is part of our engineering doctrine, as we will see later.
Growth of capability in systems science and systems thinking to enable the creation and evolution of complex systems, whether dynamic energy grids or adaptive multi-functional materials.

3.
Beyond intelligence

The emergence of Artificial Intelligence (AI) from a niche specialist topic, to the explosion of Generative AI, Agentic AI and other emerging topics like Causal and Contextual AI into the mainstream public discourse has had a remarkable impact on engineering thinking. The ability to create new text, image and video with freely available tools, that improve in quality and performance by the week, is astonishing. We have seen read-across into the production of software engineering text (code, or process deliverables, like test specifications) and the first wave of commercial tools are becoming established. The transformative nature of the landscape is moving quickly. We see a re-emergence of the debate about the philosophical definition of intelligence, questions concerning the existential threats of such developments and the need for governance of such technology. We also see a re-visit of the fundamental nature of society’s relationship with technology and its ability to take us beyond our own intelligence. With the emergence of Agentic AI, we see now the opportunity to revolutionise the way we solve problems by instructing collections of agents to solve problems without precise human planning.

Whilst the generation of credible, coherent, rich text answers to questions, production of incredible original artwork, or creation of photo-realistic video is rightly considered breathtaking, it is not the same as automatically generating the next automatic upgrade to the avionics systems of a large wide-bodied commercial aircraft. Similarly, whilst large language models and the availability of natural language interfaces to power them (‘English as the new programming language’) provide a readily accessible paradigm, we must be careful to avoid having our engineering thinking constrained by this perspective. Gen AI’s ability to generate new content beyond that imagined by us humans offers great potential, but engineering leaders must understand how to bring the benefits to engineering^[13] and to manage fundamental topics like trust, ethics and intellectual property.

The engineering lens

Engineers solve problems; they use science and mathematics to change the world. Engineers are interested in both how they engineer and what they engineer, both of which are shaped by the environment in which the engineering is taking place – the problem space. This definition sets the scene for the application of Gen AI, and in this report, we identify three aspects of this engineering lens that must be considered when applying Gen AI.

Productivity and design: Gen AI has already shown us its ability to accelerate productivity, most obviously shown with text, and increasingly with text-related deliverables, like code – all contributing to how companies engineer. They can draft outputs faster with Gen AI. But the domain is broader than simply text. Gen AI techniques can be deployed on any topic where both a grammar and a training set exist. So, engineering leaders must consider the potential for drafting deliverables in mechanical engineering, chemical engineering, structural engineering and electrical engineering. This breadth of application places demands on the underlying models, and of course around challenges on the sufficiency of the training set. Generative techniques can also be applied to what companies engineer – they can find new designs using generative technologies, new parts, new layouts, new materials, new chemicals, new designs… Gen AI can boost productivity, and it can also lead to new intellectual property. The addition of agentic thinking introduces further turbo charging, as well as opportunities for hyper-personalization and more unstructured, independently minded problem solving.
Engineering flow: Gen AI is not magic – even though at times it seems like it is. The technology is probabilistic, it will lie as frequently as it tells the truth and is heavily impacted by the biases and completeness of its training set. Engineering leaders therefore must consider how to embed the application of Gen AI into a broader process, or engineering flow. The quality controls, and the interplay between the tooling, the quality of the training set, the cost of the compute, the choice of models, the use of additional models and references – they all matter – even more so with agentic implementations, where further levels of autonomy are possible. Any application of Gen AI for engineering problem solving must be placed, and evaluated, inside the broader relevant engineering flow. AI is becoming a ‘systems design’ topic, and the combination of the technology with other engineering tools and processes is where the value acceleration will really happen.
Industrial constraints: Engineering operates inside a specific set of constraints – societal expectations on safety, reliability and risk; legal contexts and risk frameworks that are quite different the world over; and standards that govern connectivity, interference, safety, compliance, etc. Any approach to Gen AI must be anchored in the (often) domain-specific industrial contexts. Operational environments can vary – access to compute, hazardous environments, or resource constrained scenarios – all place constraints on how Gen AI can be implemented.

4.
Digital fabric

Our enterprises, and increasingly the products and services they deliver, are founded on digital technologies – a global fabric of connected communication, computation and data. These technologies are crucial for supporting the other themes in this section, by providing the necessary capabilities for real-time data processing, collaboration, coordination and system integration. This interconnected digital ecosystem is essential for ensuring the creation and scaling of engineering solutions. Engineering leaders must be cognizant of the trends and disruption that exist across all levels of this technology domain, from global information infrastructure down to fundamental physics. They should consider:

Universal connectivity.
Virtual worlds and virtual environments.
Advanced architectures in software, data, and semiconductors.
Quantum technologies.

Universal Connectivity – the new internet

Modern devices are expected to be always connected, and this has given rise to new engineering practices and technology convergence.

Today’s devices – at work and home – run intelligent applications. They connect to services running on various local compute platforms and to increasingly autonomous systems deployed and trained in the cloud. At all points, applications sense and capture their surroundings and create actionable insights.

The resources, requirements and available solutions obviously differ significantly across the architectural spectrum. Distances, latencies and data volumes will vary by orders of magnitude, as will the available resources (weight, power, size) – from lightweight devices weighing grams and using fractions of a watt, to campus-scale data centers consuming GW. Connectivity solutions will have to support all scales, from low-range connectivity for sensors to terrestrial cellular networks (like 5G, 6G), fixed networks and satellite communications. End-points will also range from hyperscale infrastructure to low-power sensors in, for example, industrial systems, vehicle-to-infrastructure communication, and infrastructure monitoring.

This diverse and pervasive combination of processing capacity and communication both demands and enables new business outcomes and models. For example, provision of functionality (data, experience) in the right place, large scale intelligence that exploits the coverage and volume of data available, and easier access to operational data through seamless connectivity. The technical convergence offered by modern networks will need to be supported by converged business models to facilitate access to a combination of connectivity, computation and data.

The network itself is a complex collection of devices; as operational data, and the models to exploit them become more widely available, it will become possible for the network to react to network events without human intervention. This gives rise to the autonomous, self-healing, self-organizing, network. The network becomes a platform that defines the ‘new internet’.

Advanced architectures

New digital capabilities depend on, and generate the opportunity for, new approaches to structuring and organizing the things that we create – new architectures. These will apply on many levels, but we can expect common architectural themes – driven by the distributed connected nature of the underlying technologies – to emerge. Consider three examples:

Composable software architectures

Traditional software architectures originate in environments where every ‘build’ scenario is costlier than taking off-the-shelf ‘buy’ solutions. This leads to architectures driven by APIs that connect pre-existing components. The increasing power of Generative AI enables new approaches, where meta-models can define much of a system, including its activities and behaviors, as well as its interfaces. Applications may be generated using technology or company-specific Gen AI, with some level of standardization and uniformity introduced by integration patterns. Architecture elements of entire systems may ultimately be generated this way.

Ontologies and knowledge
graphs

The digital fabric depends not only on physical infrastructure, but also on an ecosystem of abstractions and frameworks that define the structures and nature of our repositories and interfaces. These abstractions are changing under the influence of new technologies, like Gen AI and non-relational databases. Traditional systems are typically characterized by relational models and taxonomies that unify and generalize individual characteristics. New approaches, such as retrieval assisted generation in AI, exploit more flexible models of knowledge. Ontologies ^[9], rather than the more restricted classifications, underly new data management models.

Chiplets: a semiconductor industry breakthrough

The semiconductor industry is shifting towards chiplet architectures^[7]to address challenges in transistor scaling, voltage reduction, and limitations of external memory performance. Chiplets are specialized modular components that use advanced packaging techniques to integrate various functionalities into a unified system. This approach offers advantages, like faster time-to-market, flexible configurations, reduced risks and costs, and the integration of cutting-edge technologies. The growing demand for high-performance computing, especially in AI applications, is driving chiplet adoption across various sectors. While chiplets face challenges, like design overhead and latency, their evolution is reshaping the landscape of high-performance computing components, promising a new era of powerful and flexible integrated circuits.

Virtual worlds
and virtual environments

Virtual engineering
environments

Emerging technologies in AI, real-time 3D processing, and immersive systems are expanding digital transformation, offering innovative ways of digital interaction for various audiences. These applications, based on virtual models of physical realities, cater to specialists, industrial staff, and wider communities. They provide benefits, including increased efficiency and productivity, more personalized customer and employee relationships and enhanced user engagement via immersive experiences. These human-centric digital experiences blur the lines between physical and digital realms, supporting rapid decision-making and insight generation across organizational and client bases.

Digital world models need not have comprehensive 3D geometric models to be useful. Applications of models already exist – as virtual engineering – at several levels but, these can be expanded, their fidelity improved and their use as a means of collaboration between engineers can be developed with improving technology.

For example, virtual vehicle networks (virtual ECUs) may be used in automotive electrical/electronic/software system development. These can be hosted in a shared environment, alongside the development tools, to give a complete, scalable, globally accessible end-to-end tool chain that supports collaboration across the supply chain, while maintaining standardized governance. Test and integration activities can be started early, before representative physical hardware is available, and can exploit centralized tools, global delivery and improved utilization of capital equipment.

The industrial
metaverse

Employee and
customer experiences

The industrial metaverse ^[6]leverages these emerging technologies to build next-generation digital twins that aim to enhance operational performance in industrial settings. These digital twins are interconnected, immersive, and accessible from anywhere. This approach enables remote inspection, control, and operation of industrial equipment, as well as more general virtual design, simulation, and remote collaboration. Potential benefits include improved efficiency, faster decision-making, optimal resource use, reduced travel, and less physical testing.

More general human-machine interactions can exploit new technologies in many exchanges. For the workforce, they help to attract employees through new hiring and onboarding processes, meeting flexible working hours and remote working practices, and the key application of learning and training. Immersive learning is transforming operational, technical and soft-skill training, fostering engagement and productivity for the company’s players. For customers, these interactions have the potential to transform people's relationships with brands, products, services, or entertainment content. New products and services can be envisaged in advance through multisensory immersive experiences that surpass the level of interaction they can expect in person. Applications include product visualization, customization, digital showrooms, and social engagement, including gamification and loyalty programs.

Quantum
technologies

Quantum cryptology
and communication

In addition to the evolution of known technologies, and disruption arising from new integrations and convergence, fundamental advances in science and technology clearly have the potential to disrupt existing technologies and industries, and the rise of applications of quantum physical effects exemplifies this ^[8].
We can consider at least three application areas in engineering:

Quantum cryptology and communication.
Quantum sensing.
Quantum computation.

General applications of quantum computation will be considered below, but its application (even conceptually) to cryptology is of such import that immediate actions are justified. The significance of quantum cryptology is that many current encryption schemes would become much weaker if practical quantum computation were available. The risk to (even historical) data is sufficient that the development and adoption of quantum-safe technologies is already a necessity in critical applications.

Conversely, quantum communication (data exchange secured in part by the quantum properties of the data link) is already practical for some applications (e.g. key exchange).

Quantum
sensing

Quantum
computing

Another potential application of quantum techniques that may gain short-term traction lies in quantum sensing – using quantum effects to detect and measure a range of phenomena. Some established sensors, e.g. in photonics, depend on quantum effects, but other applications are becoming practical, particularly in imaging, and in detecting and measuring gravity. Quantum gravity sensors are already sufficiently sensitive to help map subterranean structures.

The motivation for much quantum research is the prospect of quantum computation: the use of computational elements able to address multiple quantum states, and thus break fundamental limits of classical computation and algorithms.

Current quantum computers are confined in scope to research models, limited by the scale and quality of their quantum behavior, but techniques like error suppression, error mitigation and error correction are being researched. We see an opportunity to leverage broader capabilities in data-driven research (including machine learning) alongside the specific properties of quantum systems to access unprecedented research opportunities in relevant fields. Applications can be expected in areas where complex correlated behaviors, intractable to conventional computation, are crucial. For example, complex chemistry, materials science and biological simulation.

5.
Velocity
of impact

Engineering activities are driven by, and take place within, human societies. Engineering activities are also motivated by market needs and carried out by staff and collaborators, subject to regulations and legislation set by society as a whole. So, engineers must ensure that their innovations contribute positively to society, fostering trust and acceptance, balancing the impact on social value with the need for commercial advantage.

Humans at the center of engineering

Responsible innovation
As mentioned above, engineers must ensure that their innovations contribute positively to society. Products and systems that fail to achieve this will struggle to achieve sustainable markets and may face significant legal or regulatory barriers. Expectations will vary significantly across locations and industries. Self-driving cars, railway control systems, aircraft controls and nuclear power station protection systems will face very different demands, and these demands will differ by country. The process of setting achievable and reasonable expectations, while taking benefit of technologies and commercial opportunities, is necessarily a dialogue between industry and society. Social acceptability is represented, sometimes slowly and imperfectly, by regulatory and legislative activities.

Those leading engineering programs must consider the impact of technology in the context of more comprehensive social benefits, and acceptance accelerated by appropriate stakeholder engagement. The ‘responsible innovation’ approach, for example, provides a framework and guidance for identifying and engaging with a broad stakeholder base and incorporating their input in engineering processes and regulatory evolution.

Responsible and sustainable operations
The social and environmental impacts of engineering are not confined to design and development decisions ^[14]: impacts arise from our manufacturing, delivery and operational activities ^[15]. Our social and legislative environment demands increasing transparency across our whole lifecycle, monitoring and enhancing our efficiency, sustainability, and resilience.

Sustainability here goes beyond reducing energy consumption or carbon footprints; it also encompasses social and ethical aspects. It considers environmental impacts, working conditions, supplier ethics, responsible resource management, waste reduction, and much more.

Themes explored in this document impact the development of robust and responsible operations in several ways: the introduction of additive layer manufacturing to shorten supply chains and simplify inventory management, global data platforms to optimize logistics, and bio-inspired architectures to increase product resilience, for example. These changes will not yield human or social benefits in isolation however, understanding the impact of these emergent technologies on of our operations is needed.

Resilience and security concerns
We currently face a new geopolitical environment that is more volatile than has been the case for several decades. This poses challenges for defense, and for resilience and security across all sectors. We see rapid changes in the type and scale of defense procurement, restructuring of consortia and supply chains, challenges to technology sovereignty and a consequential gap in skills and knowledge. Engineering responses to these challenges will exploit the trends discussed above, for example:

The rapid development of products and platforms to address new defense needs demands agility; achievable through the use of digital engineering and model-based development.
Restructuring of supply chains, coupled with ramp-up in production, can exploit new technology in supply chain management, alongside new materials and processes, like additive layer manufacturing.
Sovereignty issues, present throughout supply chains and technology stacks, will require new technological solutions to be developed and deployed into existing systems.
To create capacity for new programs, legacy programs will have to seek alternative models of support, increasing centralization of maintenance, and updates facilitated by digital collaboration.

In addition, demands for improved assurance of security and reliability contribute further to the challenges of compliance that engineering leaders face.

Assurance and
regulatory compliance

Trust and AI

The above factors are increasingly reflected in the requirements for these leaders to report and justify activities to independent agencies. Complying with these requirements is part of our engineering activities, and doing so efficiently becomes a key differentiator.

Advances in AI are driving significant disruption across many industries and activities, but questions about the ethics, trust and dependability of AI are as old as the concept itself. Concerns around bias and lack of transparency apply broadly across many AI applications, but in engineering contexts – as part of an engineering product or process – we have further concerns about the accuracy and correctness of AI. The fundamental technology underlying AI is the same as any software or electronic system: the difference from previous practice comes in how these systems are configured, programmed or trained. AI is characterized by statistical results, rather than the traditional logical models of software. Traditional assurance approaches, based on testing, review and inspection, are impractical for AI.

Some engineered systems may be required to have no more than one (precisely characterized) failure in a billion hours or a billion operations. Ergo, purely statistical approaches will not provide the necessary assurance. Defense-in-depth or alternative routes are being sought as verification to guarantee the behavior of these systems. These may be procedural (e.g. audits of development and training processes) or inherent in our products (e.g. a safety wrapper around an AI function). Encouraging results are emerging (see above) in Explainable AI or architectures that enforce bounded behavior. Additional constraints and checks can be applied to the validity of the results obtained, or classical control theory approaches to bounding the outputs of a system can be used.

Any single approach to this complex problem is unlikely to meet broad assurance needs. As such, those leading engineering innovation programs should expect to adopt multi-faceted assurance processes or Hybrid AI implementations.

Data-driven
regulation

Sustainability
assurance

Like AI systems, products that deal with variable real-world behavior require statistical consideration of large data volumes for regulatory acceptance. This places the generation and management of evidence at the center of the approval process.

In pharmaceuticals, for example, regulatory approval is shifting from a traditional document-centered approach to a data-driven approach using ongoing assurance of data sets that are continually reviewed and updated. This approach also allows for data sharing across markets and jurisdictions, moving away from presenting static datasets to multiple regulators separately.

In aerospace, digital techniques like models, simulations, and analysis are used to supplement physical tests. Digital certification combines dedicated verification and validation activities with extensive use of operational data and advanced analytics, potentially reducing the need for costly physical tests. While evidence quality must remain sufficient for regulatory approval, this enables new discussions about the necessary extent of physical verification and validation.

In many jurisdictions, sustainability is an increasingly regulated topic and market driver. Corresponding requirements to report non-financial performance to shareholders, customers and regulators thus must be considered by global companies. There are also growing requirements on the type and quality of information needed (for example, by the EU Corporate Sustainability Reporting Directive and proposed Green Claims Directive). Digital continuity, relevant system models and connected collaborative ways of working become key to ensuring that sufficient and accurate data is available: for product life cycle analysis, supply chain optimization, and strategic decision making. While this has some common factors with other assurance topics (eg. safety or security) the complexity and novelty of the regulation means that data may be difficult to obtain and decision criteria may not be clear. Engineering leaders can expect to see applications of AI in gathering and structuring information, and large-scale simulation employed to support decision making.