A new SimScale survey has revealed a widening divide between engineering organisations experimenting with artificial intelligence and the small group that has successfully deployed it at scale.
According to SimScale’s State of Engineering AI 2026 report, 80% of engineering organisations are running AI pilots, while only 9% have established mature, scaled AI programmes.
The findings suggest that interest in engineering AI is growing quickly, yet many companies remain unable to move beyond isolated experiments and integrate AI into everyday design, simulation and product-development workflows.
For machine builders, component manufacturers and engineering businesses, the report highlights the commercial difference between discussing AI and using it to accelerate real engineering decisions.
SimScale Survey Covers 350 Senior Engineering Leaders
The full SimScale Engineering AI report is based on responses from 350 senior engineering leaders working for organisations with more than 1,000 employees.
Participants included chief technology officers, vice presidents of engineering, heads of engineering and senior simulation leaders across the United States, United Kingdom and Germany.
Industries represented included:
- Machinery and industrial equipment
- Automotive and transportation
- Electronics and high technology
- Energy
- Architecture, engineering and construction
- Life sciences and healthcare
The research was independently administered by Global Surveyz Research during February 2026.
80% Are Experimenting, While Only 9% Have Scaled AI
The headline finding from the SimScale survey is the scale of the gap between experimentation and operational deployment.
The proportion of organisations experimenting with AI pilots has risen from 42% in 2025 to 80% in 2026. However, the percentage reporting mature, scaled AI programmes has increased only slightly, from 7% to 9%.
This means engineering teams are adopting AI at the pilot level far more quickly than they are building the infrastructure, processes and governance required to use it consistently across the organisation.
SimScale describes this position as “pilot purgatory”, where companies can demonstrate that an AI application works but struggle to connect it with established engineering systems, data, teams and decision-making processes.
The pattern reflects a wider challenge across industrial AI in manufacturing. Manufacturers increasingly have access to capable AI tools, while practical deployment depends on integration, ownership, workforce confidence and measurable business outcomes.
AI Is Already Being Used in 36% of Engineering Projects
Despite the difficulty of scaling AI across entire organisations, the technology is already influencing a significant proportion of engineering work.
The report found that an average of 36% of design and simulation projects used AI or agentic engineering during the previous 12 months.
SimScale describes this as a potential tipping point where AI begins moving from a specialist experiment into a normal part of the engineering process.
Most respondents reported using AI in between 26% and 50% of their projects, suggesting that many engineering departments are expanding deployment gradually while maintaining conventional workflows alongside the new systems.
This development connects with the rise of the manufacturing copilot, where AI assistants and agents are being integrated into CAM, simulation, robotics, maintenance and production-management software.
AI Workflows Can Triple Simulation Speed
One of the strongest findings concerns the speed of engineering simulation.
Conventional simulation requests took an average of 17 hours, compared with approximately six hours using AI-enabled workflows.
SimScale says this represents close to a threefold improvement in turnaround time. The most advanced teams reported completing some simulation requests in less than one hour.
Faster simulation can allow engineering teams to test design decisions earlier, reduce waiting between departments and keep analysis aligned with the speed of product development.
It can also expand the number of ideas that engineers are able to investigate. Teams using AI-enabled workflows evaluated an average of 56 design variants per programme, compared with 17 variants through conventional processes.
That gives engineering teams the opportunity to explore more possible designs, identify better-performing configurations and reduce the risk of reaching physical prototyping with an unsuitable concept.
Simulation is also becoming increasingly connected with digital twin technology in manufacturing, where virtual models are used to test machines, products and processes before changes are introduced in production.
Engineering AI Could Accelerate RFQ Responses
The commercial implications extend beyond product development.
According to the report, engineering teams using AI-supported workflows achieved average RFQ and technical-bid turnaround times of two days, compared with six days using conventional workflows.
In 11% of cases, AI-enabled teams completed the engineering work required for an RFQ response in less than one day.
For machine builders, system integrators and specialist component manufacturers, that could create a direct sales advantage.
Faster engineering analysis provides more time to:
- Explore alternative technical solutions
- Improve cost and performance assumptions
- Produce stronger proposals
- Respond to more opportunities
- Validate whether a proposed design is achievable
This is particularly relevant for companies producing complex, customised machinery or components where every quotation requires engineering input.
Similar pressures are influencing the development of AI CAM software, automated estimating platforms and intelligent production-planning systems.
Data Remains the Biggest Barrier
The most commonly reported barrier to scaling engineering AI was data preparation and availability, cited by 74% of respondents.
Governance and compliance concerns were identified by 48%, while 42% highlighted software interoperability.
Other barriers included:
- Lack of AI skills and knowledge
- Intellectual-property concerns
- Cultural resistance
- Computing constraints
- Legacy desktop CAE systems
However, SimScale argues that companies may be overestimating the amount of perfectly structured data required before an AI project can produce value.
Physics AI applications such as surrogate modelling can require large volumes of organised simulation data. Agentic assistants, workflow automation and engineering-support tools may be able to operate with less preparation by using existing specifications, procedures and contextual engineering information.
Cloud-based engineering platforms may help address some of these barriers by connecting simulation, computing resources, APIs and engineering data within a shared environment.
SimScale found that 75% of organisations with mature AI programmes identified cloud-native platforms as an important enabler of successful deployment.
SimScale Opens Engineering AI Agents to Enterprise Teams
The survey has been released as SimScale expands its own Engineering AI agents to enterprise customers.
The agents are designed to extract technical intent from project documents and orchestrate simulation workflows covering:
- CAD preparation
- Meshing
- Solver configuration
- Parallel simulation
- Validation
- Reporting
The system can use an organisation’s engineering standards, preferred simulation methods and compliance rules to help automate work normally requiring specialist input.
SimScale says its agents can process RFQ documents, inspect CAD geometry, identify missing information and execute multi-stage simulation workflows while maintaining engineering oversight.
This places SimScale within a growing area of software, CAM and industrial AI where AI is moving from passive recommendations toward the execution of defined engineering tasks.
99% Expect Meaningful Business Value From AI
Confidence in the commercial potential of engineering AI is now extremely high.
The report found that 99% of respondents expect their organisation to realise meaningful business value from AI or agentic engineering during the next 12 months.
Of those surveyed, 24% said they were very confident and 75% said they were somewhat confident. Only 1% lacked confidence.
David Heiny, co-founder and CEO of SimScale, said:
“What we’re seeing now is a shift from experimentation to scaled execution.”
Heiny said the rapid expansion of pilots and mature programmes indicates a clear wave of AI adoption across engineering.
Human Oversight Remains Central
Engineering organisations are allowing AI to participate in important decisions, although most continue to require human review.
The report found that 87% of respondents permit AI to contribute to pass-or-fail decisions at design gates under defined governance arrangements.
Within that group, 79% allow AI to participate with formal review points, while 8% permit autonomous decisions without an additional approval gate.
A total of 71% require mandatory human review of every AI output.
This indicates that engineering AI is being introduced through controlled systems in which engineers remain accountable for verifying results and approving consequential decisions.
MTN Analysis
The most important result from the SimScale survey is the commercial separation forming between companies that experiment with AI and those that operationalise it.
The organisations achieving value have established ownership, governance, cloud infrastructure and defined objectives for moving pilots into production.
For machine builders and metal manufacturers, this is a critical point. A successful AI deployment needs to improve an identifiable part of the engineering or production process.
That could mean:
- Completing simulations more quickly
- Testing more design alternatives
- Reducing RFQ response times
- Improving first-time-right engineering
- Automating repetitive setup work
- Making specialist knowledge available across a wider team
Manufacturers evaluating AI at IMTS 2026 and other major industrial exhibitions should ask suppliers for evidence of how pilot projects are converted into repeatable production workflows.
The technology is progressing rapidly. The competitive advantage will increasingly come from an organisation’s ability to deploy it consistently, measure its results and integrate it into existing engineering operations.
FAQ
What is the State of Engineering AI 2026 report?
The State of Engineering AI 2026 is a SimScale survey of 350 senior engineering leaders examining AI adoption across product design, engineering and simulation workflows.
How many engineering organisations have mature AI programmes?
Only 9% of organisations surveyed reported having mature, scaled AI programmes, while 80% were experimenting with pilot projects.
What does AI pilot purgatory mean?
AI pilot purgatory describes the stage where an organisation has tested an AI application successfully but has not integrated it into normal, repeatable business or engineering operations.
How much engineering work currently uses AI?
The report found that an average of 36% of design and simulation projects used AI or agentic engineering during the previous 12 months.
Can AI improve engineering simulation speed?
Yes. SimScale found that AI-enabled simulation requests took an average of six hours, compared with 17 hours through conventional workflows.
What prevents engineering companies from scaling AI?
The main reported barriers were data preparation, governance, compliance, software interoperability, skills shortages and legacy engineering systems.
Will AI replace simulation engineers?
The survey suggests that most organisations are using AI within defined governance systems. Human engineers continue to review results, approve decisions and maintain technical accountability.
Further Reading
Is 2026 the Rise of the Manufacturing Copilot?
Digital Twin in Manufacturing 2026: The Plain-English Guide for Factories
AI CAM Software 2026: hyperMILL vs NX vs Mastercam vs Fusion
IMTS 2026 Industrial AI: How AI Is Moving Into Job Shops, CAM and Quality Control
External Sources
Read the State of Engineering AI 2026 report
Download the full SimScale survey report
Read SimScale’s official report announcement
Learn about SimScale Engineering AI agents
