Safety-critical software

Testing and investigation
for safety-critical software

For embedded and complex systems where failures need to be explainable, reproducible and traceable.

Midair connects static analysis, controlled test runs and runtime evidence, so critical software behavior can be investigated in one context.

Critical failures are rarely explained by one signal.

In safety-critical systems, a defect may depend on source code, timing, hardware state, configuration or a specific sequence of events.

Finding that something failed is only the beginning. The harder part is understanding exactly why it happened and proving that the changed system behaves differently afterwards.

Code-level risks

Unsafe behavior may originate in execution paths or project-specific rules before the software ever runs.

Environment-dependent failures

Some defects appear only on particular hardware, configurations or operating conditions.

Rare sequences matter

A failure may require a specific order of events that is difficult to recreate manually.

Evidence must survive the fix

A corrected build is useful only if it can be compared with the conditions and evidence behind the original failure.

Midair

Build an evidence trail across the verification workflow.

Before execution

Analyze risky behavior

Visao examines execution paths and project-specific rules to identify defects and unsafe behavior before the software runs.

During verification

Run controlled scenarios

TS Factory coordinates repeatable tests across devices and machines while keeping the environment behind each run.

During execution

Capture runtime evidence

Delta collects logs, system events and resource data while the system is operating.

Midair keeps the findings, test conditions and runtime evidence connected, so the original failure and the verified change can be reviewed in the same context.

What safety-critical software verification can cover

Critical code paths

Analyze behavior where incorrect execution can affect the wider system.

Hardware-dependent behavior

Test software whose result depends on devices, controllers, interfaces or physical state.

Fault and recovery scenarios

Reproduce behavior during failure, restart, degraded operation or recovery sequences.

Release and regression verification

Compare critical behavior before and after software or configuration changes.

Distributed system behavior

Coordinate scenarios where several devices, hosts or services influence the same result.

Long-running conditions

Investigate failures that only appear after repeated operations or extended execution.

Traceability means keeping the technical evidence connected.

A safety-critical investigation often crosses several layers of the system.

Midair keeps the technical context behind a result connected, including:

•

Source finding

What code behavior or rule triggered the investigation.

•

Software version

Which build was running.

•

System configuration

Which device, hardware or environment produced the result.

•

Test run

Which scenario and inputs were executed.

•

Runtime evidence

What the system actually did while running.

•

Verification run

Whether the same scenario changed after the fix.

Verification is more than repeating a test.

When the consequences of failure are serious, a green result without context is weak evidence.

The important question is whether the same relevant conditions can be reconstructed and whether the changed system behaves differently under them.

Midair is designed to preserve that context across analysis, testing and runtime investigation, so each result stays tied to the others.

Where this becomes useful

Embedded control systems

Software where behavior depends on hardware state, timing and physical interfaces.

Industrial systems

Controllers and distributed environments where incorrect behavior can interrupt or damage operations.

Network infrastructure

Firmware and system software where failures can affect many connected devices or services.

High-reliability devices

Products where unexpected software behavior can create significant operational or user impact.

Complex regulated environments

Projects that require stronger verification discipline and documented engineering evidence.

Safety-critical workflows and industry standards

Midair supports engineering verification. It does not certify a system by itself.

Safety-critical development may be governed by standards such as ISO 26262, IEC 61508, IEC 62304 or DO-178C, depending on the industry and system.

Those standards define broader lifecycle, process, documentation and assurance requirements. A testing or analysis platform alone does not make a system compliant or certified.

Midair can support the technical verification workflow by keeping analysis findings, test runs, configurations and runtime evidence connected.

Static analysis matters more when generic rules are not enough.

Safety-critical code often contains project-specific APIs, state machines and operation sequences that generic defect rules do not understand.

Visao supports custom detectors and behavior-oriented analysis, making it possible to encode project-specific constraints into repeatable checks. Interpretica's static-analysis research positions Visao around execution paths, program states and custom detectors. Broad compliance reporting is outside that focus.

Start with one critical failure or verification scenario.

Bring one codebase, one target environment and one behavior that is difficult to explain or verify reliably.

Connect the relevant analysis, test run and runtime evidence in Midair and evaluate the complete investigation workflow on a scoped pilot.