Automotive safety development is moving further into the virtual world as automakers seek faster ways to evaluate crash performance before building physical prototypes. Autoliv is expanding that shift with its Human Body Model (HBM) Safety Suite, a virtual safety platform that gives engineers a more detailed digital view of how occupants may respond during crashes. Toyota is the first customer to begin evaluating the technology.

Autoliv Takes Human Crash Testing Further Into the Virtual World

Physical crash tests remain a critical part of vehicle safety validation, but they have an obvious limitation: engineers cannot easily reproduce every crash configuration, occupant profile or injury mechanism in the real world.

That is creating space for virtual testing.

Autoliv has announced its Human Body Model (HBM) Safety Suite, an integrated virtual safety platform designed to simulate occupant behavior during crashes and help automotive engineers evaluate safety performance earlier in vehicle development.

Toyota has already started using the platform to evaluate its capabilities as part of its virtual safety development activities, according to Autoliv. The safety supplier plans to make the platform more broadly available during 2026.

The technology represents a broader transition in automotive engineering: instead of relying on physical prototypes and crash dummies for every development decision, automakers can increasingly combine physical validation with computer-based simulations.

A digital human body for crash simulation

At the center of Autoliv’s platform is a Human Body Model, a detailed digital representation of human anatomy designed to provide insights into occupant response and potential injury mechanisms that conventional physical crash dummies cannot capture.

Autoliv’s HBM Safety Suite combines the model with software for simulation, analysis, visualization and decision-making. The company’s SAFER HBM can also be adapted to represent road users with different sex, age, height and weight characteristics.

That flexibility is increasingly important as automakers look beyond a single standardized occupant.

A physical dummy can provide highly valuable measurements, but a digital human model can expose additional information about how different parts of the body move, deform and experience loads during a collision.

The objective is not to eliminate physical crash testing. Instead, virtual testing can expand the number of scenarios engineers can investigate before committing to physical builds.

Virtual testing is becoming part of mainstream vehicle development

The automotive industry has already been using simulation extensively for areas such as vehicle dynamics, aerodynamics, structural analysis and powertrain development.

Crash safety is now becoming another major area for virtual engineering.

McKinsey estimates that China’s automotive manufacturers use software simulation and virtual prototypes for around 65% of their testing, compared with approximately 40% to 50% in other regions. The consultancy estimates that maximizing virtual testing could cut the number of physical prototypes required by half and potentially reduce time to market by nine to 11 months.

That competitive pressure is significant.

Vehicle programs are becoming more software-intensive while automakers simultaneously face shorter development cycles and increasingly complex safety requirements. Faster simulation allows engineering teams to test more alternatives earlier, rather than waiting until physical prototypes are available.

For safety engineers, that can mean evaluating restraint systems, occupant positioning and crash configurations earlier in the design process.

Euro NCAP is pushing the industry toward human-body simulation

The regulatory and consumer-testing environment is also beginning to reinforce the shift.

Euro NCAP’s virtual-testing work includes Human Body Model simulations, and its protocol documentation states that HBM simulations are planned to become part of Euro NCAP assessments in 2029. The organization currently continues to use physical dummies as standardized assessment tools while HBM methodologies are being developed and validated.

Euro NCAP’s protocol already provides incentives around virtual testing. Its crash-protection documentation states that vehicle manufacturers providing HBM simulation data can qualify for full points for certain virtual load cases, while manufacturers that do not provide the data face a reduced virtual score.

That makes virtual safety engineering more than a technology experiment.

As testing organizations increasingly incorporate simulation into assessment frameworks, automakers and suppliers need validated digital models and workflows capable of producing reliable results.

Autoliv’s strategy is consequently aimed at a market where virtual crash simulation is likely to become a standard engineering capability rather than an optional research tool.

The AI connection is bigger than the human model

Although the HBM Safety Suite is fundamentally a simulation platform, its development fits into the wider AI transformation of automotive engineering.

Modern vehicle development increasingly relies on machine learning, surrogate models, automated scenario generation and AI-assisted engineering to reduce the amount of physical testing required.

McKinsey has noted that deep-learning surrogate models can replace some physical tests with faster virtual assessments, while applied AI can generate large numbers of additional scenarios for vehicle testing.

The implication is that automotive simulation is becoming a computational pipeline rather than a single engineering calculation.

Engineers can create a digital vehicle, introduce different crash conditions, simulate occupant responses, analyze injury mechanisms and compare alternative safety-system configurations—all before running the corresponding physical test.

That does not make the virtual result automatically correct. Simulation quality depends heavily on model validation, input data, solver accuracy and accurate representation of vehicle structures and restraint systems.

Euro NCAP itself highlights these limitations. Its HBM documentation notes that current Human Body Models are not yet standardized for all injury predictions and that simulation results can be affected by variability in the HBM as well as limitations in modeling vehicle interiors, airbags and other components.

That is why the combination of virtual and physical testing remains important.

From crash dummies to digital occupants

Autoliv’s platform could ultimately change how engineers think about occupant safety.

Traditional crash testing largely revolves around standardized physical representations of occupants. Human Body Models offer the possibility of evaluating a broader range of body characteristics and understanding injury mechanisms at a much more granular level.

That could become increasingly relevant as automotive safety moves toward more personalized protection.

Autoliv already describes adaptive safety as a direction in which restraint systems dynamically respond to characteristics such as occupant size, weight, age and seating position.

A detailed digital human model can potentially provide the simulation foundation needed to develop and validate those systems.

For automakers, the value is not simply fewer crash tests. It is the ability to explore more safety configurations, identify problems earlier and make engineering decisions with greater visibility into how different occupants may respond.

Competition is forming around simulation ecosystems

Autoliv is entering an increasingly sophisticated ecosystem that includes engineering simulation providers, digital-twin platforms, AI infrastructure companies and automotive technology suppliers.

Companies such as Ansys, Siemens, Dassault Systèmes and Altair have long provided simulation and digital-engineering technologies, while automotive OEMs and suppliers increasingly develop proprietary models and virtual-validation environments.

The competitive advantage may therefore shift toward platforms that combine validated human models, vehicle simulation, analysis tools and engineering workflows rather than standalone simulation software.

Autoliv has one potential advantage: its position as a major automotive safety supplier gives it direct domain expertise in airbags, seatbelts, restraint systems and occupant protection.

Its HBM Safety Suite builds on that safety knowledge while adding a software-driven development layer.

The road to virtual-first safety engineering

The automotive industry is unlikely to abandon physical crash testing anytime soon.

Instead, the development process is becoming simulation-first and validation-driven.

Engineers can use virtual testing to explore thousands of possible conditions, narrow down promising designs and identify potential failure modes. Physical testing can then validate the most important scenarios and confirm that the digital models accurately represent real-world behavior.

That hybrid model could allow automakers to develop safer vehicles while reducing development cycles and physical prototype requirements.

Autoliv’s HBM Safety Suite is therefore part of a larger transformation in automotive engineering: the vehicle is increasingly being developed twice—once as a physical machine and once as a detailed computational model.

The more accurate that digital counterpart becomes, the earlier engineers can identify safety problems and the more efficiently they can design around them.

Market Landscape

Virtual engineering is becoming a competitive factor in automotive development as OEMs seek to shorten development cycles while managing increasingly complex software, electronics and safety requirements.

McKinsey reports that leading Chinese OEMs already conduct around 65% of testing through simulation and virtual prototypes, compared with 40% to 50% in other regions.

Key market trends include:

• Human Body Models: Digital occupants that provide more detailed information about injury mechanisms than conventional dummies.
• AI-assisted simulation: Machine learning and surrogate models accelerating computationally intensive engineering workflows.
• Digital twins: Virtual representations of vehicles and components used for continuous testing and optimization.
• Scenario generation: AI enabling engineers to evaluate far larger numbers of potential crash and driving conditions.
• Virtual homologation: Increasing use of simulation within regulatory and consumer safety assessment processes.
• Hybrid validation: Virtual testing complementing rather than immediately replacing physical crash tests.

Euro NCAP’s roadmap is particularly important because its planned 2029 integration of HBM simulations into assessments could encourage wider adoption across vehicle manufacturers.

Top Insights

• Autoliv’s HBM Safety Suite uses a digital human model to provide deeper insight into occupant movement and potential injury mechanisms during simulated crashes.
• Toyota is the first customer evaluating the platform, while Autoliv plans broader industry availability during 2026.
• Euro NCAP is preparing to incorporate Human Body Model simulations into its assessment framework, with HBM testing planned for 2029.
• Virtual testing can expand the number of crash scenarios engineers investigate while reducing dependence on expensive physical prototypes during early development.
• The emerging model combines AI, simulation and physical validation rather than attempting to eliminate real-world crash testing altogether.

Power Tomorrow’s Intelligence — Build It with TechEdgeAI