1. Requirements

1.1 Notes

NPR 7150.2, NASA Software Engineering Requirements, does not include any notes for this requirement.

1.2 History

1.3 Applicability Across Classes

2. Rationale

It is important to determine the safety criticality of each software component to identify the most critical software system components and to ensure that the software safety-critical requirements and processes are followed.

3. Guidance

Classifying software essentially provides pre-tailoring of software engineering requirements, software safety requirements, software assurance requirements, and other software requirements for different types and levels of software. Beyond classification, each project evaluates the project software to determine if the software is safety-critical. 

Safety-critical is a term “describing any condition, event, operation, process, equipment, or system that could cause or lead to severe injury, major damage, or mission failure if performed or built improperly, or allowed to remain uncorrected.” 

Software safety is defined as “the aspects of software engineering and software assurance that provide a systematic approach to identifying, analyzing, tracking, mitigating, and controlling hazards and hazardous functions of a system where software may contribute either to the hazard or to its mitigation or control, to ensure safe operation of the system.” 

The project can use NASA-STD-8739.8 

Software safety criticality is initially determined in the formulation phase using the NASA Software Assurance Standard, NASA-STD-8739.8.   As the software is developed or changed and the computer software configuration items (CSCI), models, and simulations are identified, the safety-critical software determination can be reassessed and applied at lower levels. The software safety assessment and planning are performed for each software acquisition, development, and maintenance activity, and changes to legacy\heritage systems. When the software in a system or subsystem is found to be safety-critical, additional requirements in NASA-STD-8739.8, the NASA Software Assurance and  Software Safety Standard will augment those associated with the software class requirements found in this document.

See the software assurance tab for Considerations when identifying software subsystem hazard causes and for Considerations when identifying software causes in general software-centric hazard analysis.

Additional guidance related to software safety may be found in the following related requirements in this Handbook:

4. Small Projects

No additional guidance is available for small projects.

5. Resources

5.1 References

5.2 Tools

6. Lessons Learned

6.1 NASA Lessons Learned

No Lessons Learned have currently been identified for this requirement.

6.2 Other Lessons Learned

No other Lessons Learned have currently been identified for this requirement.

7. Software Assurance

7.1 Tasking for Software Assurance

1. Confirm that the hazard reports or safety data packages contain all known software contributions or events where software; either by its action, inaction, or incorrect action, lead to a hazard.
2. Assess that the hazard reports identify the software components associated with the system hazards per the criteria defined in NASA-STD- 8739.8 Appendix A.
3. Assess that hazard analyses, including hazard reports, identify the software components associated with the system hazards.
4. Confirm that the traceability between software requirements and hazards with software contributions exists.
5. Develop and maintain a software safety analysis throughout the software development life-cycle.

7.2 Software Assurance Products

• Hazard Analyses
• Assessment of Hazard Analyses and Reports
• The list of all software safety-critical components that have been identified by the system hazard analysis.
• Software Safety Analysis results, including any risks or issues.
• Analysis of updated hazard reports at review points.
• Record of software safety determination for the software project.

  
  

  Note
  title Objective Evidence
  Evidence of confirmation of traceability between software requirements and hazards with software contributions. Evidence that confirmation of all safety contributions or events in hazard reports has occurred.

  Expand
  title Definition of objective evidence
  Include Page SITE:Definition of Objective Evidence SITE:Definition of Objective Evidence

7.3 Metrics

• # of software work product Non-Conformances identified by life-cycle phase over time
• # of safety-related requirement issues (Open, Closed) over time.
• # of safety-related non-conformances identified by life-cycle phase over time.
• # of Hazards containing software that has been successfully tested vs. total # of Hazards containing software

7.4 Guidance

Safety-Critical: A term describing any condition, event, operation, process, equipment, or system that could cause or lead to severe injury, major damage, or mission failure if performed or built improperly, or allowed to remain uncorrected. (Source NPR 8715.3)

Safety-Critical Software: Software is classified as safety-critical if it meets at least one of the following criteria:

a. Causes or contributes to a system hazardous condition or event,

b. Provides control or mitigation for a system hazardous condition or event,

c. Controls safety-critical functions,

d. Mitigates damage if a hazardous condition/event occurs,

e. Detects, reports, and takes corrective action if the system reaches a potentially hazardous state.

Safety-Critical Software can cause, contribute to, or mitigate human safety hazards or damage facilities. Safety-critical software is identified based on the results of the hazard analysis and the results of the Orbital Debris Assessment Report/End-Of-Mission Plan (where applicable). Examples of safety-critical software can be found in all types of systems, including Flight, Ground Support systems, Mission Operations Support Systems, and Test Facilities. See Appendix A for guidelines associated with addressing software in hazard definitions. Consideration for other independent means of protection (software, hardware, barriers, or administrative) should be a part of the system hazard definition process.

Task 1: Confirm that the hazard reports or safety data packages contain all known software contributions or events where software; either by its action, inaction, or incorrect action, lead to a hazard.

It is necessary for software assurance and software safety personnel to begin examining possible software hazards and determining whether the software might be safety-critical as early as possible. Several steps are important in determining this:

A. Learn who the key project personnel are and begin establishing a good working relationship with them. In particular, systems analysts, systems safety personnel, requirements development personnel, end-users, and those establishing operational concepts are some of the key people with initial knowledge on the project.

B. Gather all of the initial documents listed in the requirement as well as any others that the project is developing that may contain critical information on the project being developed. Don’t wait for signature copies, but begin getting acquainted with them as early as possible. Through the working relationships established, stay informed about the types of updates that are being made as the system concepts continue to be refined. Keep a list of the documents collected and their version dates as the system matures. Potential documents that may contain critical information include:

• • Preliminary System-Level Preliminary Hazard Analyses (PHA)
  • Concept of Operations (ConOps)
  • Generic Hazard Lists (e.g. for project type, for software, or just generic hazards)
  • Critical Item List(s) (CIL)
  • Preliminary System Reliability Assessment or Analyses
  • Project/System Risk Assessments
  • Request for Proposals (RFP)
  • Computing System Safety Analyses
  • Software Security Assessment (NPR 7150.2, SWE-156, SWE-154)
  • Science Requirements Document

C. Become familiar with the documents in Step 2 and pay particular attention to the risks and potential hazards that might be mentioned in these documents. While reviewing these risks and potential hazards, think about the ways that the software might be involved in these risks. Possible example software contributions to potential hazards are found in the section below titled: Software Contributions to Hazards, Software in system hazard analysis

D. As the initial hazard analyses are being done, software assurance and software safety people confirm that these analyses are as complete as possible for the stage of the project.

Task 2: Assess that the hazard reports identify the software components associated with the system hazards per the criteria defined in NASA-STD- 8739.8 Appendix A.

Review each hazard report to see that the software components associated with the system hazards are identified, using the criteria defined in NASA-STD- 8739.8 Appendix A. The Hazard Analysis done at this point should identify the initial set of planned safety-critical components. A list of all the safety-critical components should be included in the hazard reports. Keep this safety-critical components list for Task 4 and 5.

Task 3: Analyze the updated hazard reports and design at review points to determine if any newly identified software components are safety-critical.

At each milestone or review point, review any updated hazard analyses or new hazard reports. Review the current design to determine whether any new software has been identified as safety-critical. By this point in the project, some of the software may have been identified as control or mitigation software for one of the previously identified hazards and it may not have been thought about in earlier hazard reports. Verify that this newly identified software is now included in a hazard report, is included on the safety-critical components list, and has a corresponding requirement. As the project continues and requirements mature, any newly identified safety-critical software should be added to the hazard reports, so the reports contain a complete record of all safety-critical components.

Task 4: Confirm that the traceability between the software requirements and the hazards with software contributions exists.

As the project progresses, review the hazard reports with software contributions and confirm that the associated safety-critical component is listed in the hazard reports and can be traced back to a requirement in the requirements document. Confirm these requirements trace to one or more tests, and that they include testing of the software-critical capabilities required.

Task 5: Develop and maintain a software safety analysis throughout the software development lifecycle.

Throughout the software development, starting during the requirements phase, develop a software safety analysis. Topic 8.9 Software Safety Analysis guides on doing a software safety analysis.

Software Contributions to Hazards, Software in system hazard analysis:

Hazard Analysis must consider the software’s ability, by design, to cause or control a given hazard. It is a best practice to include the software within the system hazard analysis. The general hazard analysis must consider software common-mode failures that can occur in instances of redundant flight computers running the same software. A common mode failure is a specific type of common cause failure where several subsystems fail in the same way for the same reason. The failures may occur at different times and the common cause could be a design defect or a repeated event.

Software Safety Analysis supplements the system hazard analysis by assessing the software performing critical functions serving as a hazard cause or control. The review assures the following:

1) Compliance with the levied functional software requirements, including SWE-134, 

2) That the software shouldn’t violate the independence of hazard inhibits, and

3) That the software shouldn’t violate the independence of hardware redundancy.

The Software Safety Analysis should follow the phased hazard analysis process. A typical Software Safety Analysis process begins by identifying the must work and must not work functions in Phase 1 hazard reports. The system hazard analysis and software safety analysis process should assess each function, between Phase 1 and 2 hazard analysis, for compliance with the levied functional software requirements, including SWE-134. For example, Solar Array deployment (must work function) software should place deployment effectors in the powered off state when it boots up and requires to initialize and execute (arm and fire) commands in the correct order within 4 CPU cycles before removing a deployment inhibit. The analysis also assesses the channelization of the communication paths between the inputs/sensors and the effectors to assure there is no violation of fault tolerance by routing a redundant communication path through a single component. The system hazard analysis and software safety analysis also assure the redundancy management performed by the software supports fault tolerance requirements. For example, software can’t trigger a critical sequence in a single fault-tolerant manner using single sensor input. Considering how software can trigger a critical sequence is required for the design of triggering events such as payload separation, tripping FDIR responses that turn off critical subsystems, failover to redundant components, and providing closed-loop control of critical functions such as propellant tank pressurization.

The design analysis portion of software safety analysis should be completed by Phase 2 safety reviews. At this point, the software safety analysis supports a requirements gap analysis to identify any gaps (SWE-184) and ensure the risk and control strategy documented in hazard reports are correct as stated. Between Phase 2 and 3 safety reviews, the system hazard analysis and software safety analysis supports the analysis of test plans to assure adequate off-nominal scenarios (SWE-062, SWE-065 a). Finally, in Phase 3, the system hazards analysis must verify the final implementation and verification upholds the analysis by ensuring test results permit closure of hazard verifications (SWE-068) and that the final hazardous commands support the single command and multi-step command needs and finalized pre-requisite checks are in place.

The following sections include useful considerations and examples of software causes and controls:

Considerations when identifying software subsystem hazard causes: (This information is also included in Appendix A.1.4 of NASA-STD-8739.8)

1. 1. Does software control any of the safety-critical hardware?
   2. Does software perform critical reconfiguration of the system during the mission?
   3. Does the software perform redundancy management for safety-critical hardware?
   4. Does the software determine when to perform a critical action?
   5. Does the software trigger logic to meet failure tolerance requirements?
   6. Does the software monitor hazard inhibits, safety-critical hardware/software, or issue a caution and warning alarm used to perform an operational control?
   7. Does the software process or display data used to make safety-critical decisions?
   8. Does the flight or ground software manipulate hazardous system effectors during prelaunch checkouts or terminal count?
   9. Does the software perform analysis that impacts automatic or manual hazardous operations?
   10. Does the software serve as an interlock preventing unsafe actions?
   11. Does the software contain stored command sequences that remove multiple inhibits from a hazard?
   12. Does the software initiate any stored command sequences, associated with a safety-critical activity, and if so, are they protected?
   13. Does software violate any hazard inhibits or hardware redundancy independence (channelized communication/power paths, stored command sequences/scripts, FDIR false positive, etc.)?
   14. Can the software controls introduce new hazard causes?
   15. Are the software safety-critical controls truly independent?
   16. Can common cause faults affect the software controls?
   17. Can any of the software controls used in operational scenarios cause a system hazard?
   18. Does the software control switch-over to a backup system if a failure occurs in a primary system?
   19. Is the software process sensor data used to make safety-critical decisions fault-tolerant?
   20. Does the software provide an approach for recovery if the system monitoring functions fail?
   21. Does the software allow the operators to disable safety-critical controls unintentionally?
   22. Does the software provide safety-critical cautions and warnings?
   23. Is the software capable of diagnosing and fixing safety-critical faults that might occur in operations?
   24. Does the software provide the health and status of safety-critical functions?
   25. Does the software process safety-critical commands (including autonomous commanding)?
   26. Can the software providing full or partial verification or validation of safety-critical systems generate a hazard if the software has a defect, fault, or error?
   27. Can a defect, fault, or error in the software used to process data or analyze trends that lead to safety decisions cause a system hazard?
   28. Do software capabilities exist to handle the potential use cases and planned operations throughout all phases of use, and through transitions between those phases/states?

Considerations when identifying software causes in a general software-centric hazard analysis:

• Computer Reset
  • Causes
    • Reset with no restart
    • Reset during program upload (PROM corruption)
    • Boot PROM corruption preventing reset
    • Watchdog active during reboot causing infinite boot loop
  • Controls
    • Disable FDIR and watchdogs during boot
    • Redundant computers
• Memory
  • Causes
    • Memory corruption
    • Out of Memory
    • Buffer overrun
    • Deadlock (trying to write to the same memory at the same time or trying to update while reading it)
    • Single Event Upset/bit flip
    • Shared Memory (can defeat redundancy independence)
  • Controls
    • No Dynamic allocation of memory after initialization
    • EDAC/Memory scrubbing
    • Memory margin
• CPU
  • Causes
    • Cycle overrun
    • Cycle jitter
    • Stack processing/ordering incorrect (e.g., FIFO vs. LIFO vs. unknown)
    • One task failing to release shared resource
    • Single Event Upset
    • Hardware failure
  • Controls
    • CPU utilization and cycle time margin
    • Self-checking CPU cores
• Commanding
  • Causes
    • Inadvertent/Erroneous commanding
    • Failure to command
    • Command out of sequence
    • Command corrupted or illegitimate command
  • Controls
    • Command validation (cyclic redundancy check, timestamp, destination/source)
    • Acknowledgment/ negative acknowledgment
• Communication and Input/Output
  • Causes
    • Communication bus overload
    • Lack of or intermittent communication
    • Complex configuration input
    • Babbling Node
  • Controls
    • Communication bus utilization margin
    • Use of lossless protocol
    • Downstream input voting
    • External safing function
    • Bus Guardian
• Display Data
  • Causes
    • Incorrect Data (unit conversion, incorrect variable type, etc.)
    • Stale Data
  • Controls
    • Visual indication of stale data
    • Watchdog timer
• Events and Actions
  • Causes
    • Out-of-sequence event protection
    • Multiple events/actions trigger simultaneously (when not expected)
    • Error/Exception handling missing or incomplete
    • Inadvertent mode transition
  • Controls
    • Fault Management
    • Pre-requisite logic
    • Interlocks
• Timekeeping
  • Causes
    • Time runs fast/slow
    • Time skips (e.g., Global Positioning System time correction)
    • Time sync across components
    • Oscillator Drift
  • Controls
    • Diverse/redundant time sources with fault down logic
    • Robust time sync design that can deal with the loss of external time sources
    • Pre-launch checkout of Oscillators
• Timing Problems
  • Causes
    • Priority inversion
    • Failure to terminate/complete process in a given time
    • Data latency/sampling rate too slow
    • Race Conditions
    • Non-determinism
    • Static and Dynamic Analysis Tools
    • Coding standards
  • Controls
• Coding, Logic, and Algorithm failures
  • Causes
    • Division by zero
    • Bad data in = bad data out (no parameter range & boundary checking)
    • Dead code
    • Unused code
    • Non-functional loops
    • Endless do loops
    • Incorrect passes (too many or too few or not at the correct time)
    • Incorrect “if-then” and incorrect “else.”
    • Too many or too few parameters for the called function
    • Case/type mismatch
    • Precision mismatch
    • Rounding or truncation fault
    • Resource contention (e.g., thrashing: two or more processes accessing a shared resource)
    • Bad configuration data/no checks on external input files and data
    • Inappropriate equation
    • Undefined or non-initialized data
    • Limit ranges
    • Relationship logic for interdependent limits
    • Overflow or underflow in the calculation
  • Controls
    • Use of industry-accepted coding standard
    • Use of safe math libraries
    • The robust software development, quality, and safety processes
• Input Failures
  • Causes
    • Noise
    • Sensors or actuators stuck at some value (all zeros, all ones, some other value)
    • A value above or below range
    • Value in range but incorrect
    • Physical units incorrect
    • Inadequate data sampling rate
  • Controls
    • Sensor Health Checks
    • Input Validation, Sanity Checks
• User Interface Errors
  • Causes
    • Wrong commands are given by the operator
    • No commands are given by the operator
    • Status and messages not provided for operations, systems, and inhibits
    • Ambiguous or incorrect messages
    • User display locks up/fails
  • Controls
    • Two-step commands
    • GUI style guide
    • Software interlocks to prevent human error
• Configuration Management
  • Causes
    • Incorrect version loaded
    • Incorrect configuration values
  • Controls
    • Version CRC check after software/configuration load
• Security
  • Causes
    • Denial/Interruption of Service
    • Spoofed/Jammed inputs
    • An unauthorized input/access
    • Message filtering to detect spoofing
    • Ensure software has data source validation checking features
  • Controls
• PLC Processor Fault causes

Some safety-critical aspects are addressed in hardware, for example, valves failing closed when a fault occurs.

Safety products (including hazard reports, responses to launch site requirements, preliminary hazard analyses, etc.) begin with the PHA, evolving and expanding throughout the project lifecycle.