Compatibility of STPA with GM System Safety Engineering Process. Padma Sundaram Dave Hartfelder

Similar documents
Application of STPA to a Shift by Wire System (GM-MIT Research Project)

STPA in Automotive Domain Advanced Tutorial

ENGINEERING FOR HUMANS STPA ANALYSIS OF AN AUTOMATED PARKING SYSTEM

2015 STPA Conference. A s t u d y o n t h e f u s i o n o f S T P A a n d N i s s a n ' s S y s t e m s E n g i n e e r i n g

Can STPA contribute to identify hazards of different natures and improve safety of automated vehicles?

STPA based Method to Identify and Control Software Feature Interactions. John Thomas Dajiang Suo

Integrating State Machine Analysis with STPA

Functional Safety Analysis of Automated Vehicle Lane Centering Control Systems. Volpe The National Transportation Systems Center

SAFETY AND RELIABILITY ANALYSIS OF ELECTRIC POWER STEERING SYSTEM USED IN AUTOMOBILES

OPTIMORE - Optimised Modular Range Extender for every day customer usage AVL SCHRICK project summary

Introduction to Requirement Management for Safety-Critical Embedded Vehicle Systems

GM Presentation for Introducing

Potential Electronic Causes of Unintended Acceleration

Team Aware Perception System using Stereo Vision and Radar

Technical Article. ISO26262: ams deploys unique technology to meet every new safety requirement. Roland Einspieler

Analyzing Feature Interactions in Automobiles. John Thomas, Ph.D. Seth Placke

An approach based on Engineering a Safer World Systems Thinking Applied to Safety Leveson (2011)

Applying STPA to Automo0ve Adap0ve Cruise Control System. Dr. Qi Van Eikema Hommes April 18, 2012

The MathWorks Crossover to Model-Based Design

F/A-18A/B/C/D Flight Control Computer Software Upgrade

Adaptive Cruise Control System Overview

SMSIL-4THGU-E ISSUE: 12/12 SIL-SAFETY MANUAL. Upgrade Series Actuators

Our Approach to Automated Driving System Safety. February 2019

Multi-ECU HiL-Systems for Virtual Characteristic Rating of Vehicle Dynamics Control Systems

EMC System Engineering of the Hybrid Vehicle Electric Motor and Battery Pack

CASCAD. (Causal Analysis using STAMP for Connected and Automated Driving) Stephanie Alvarez, Yves Page & Franck Guarnieri

Items to specify: 4. Motor Speed Control. Head Unit. Radar. Steering Wheel Angle. ego vehicle speed control

Switching Control for Smooth Mode Changes in Hybrid Electric Vehicles

Systems-Theoretic Process Analysis: AUTOMOBILE FEATURES FOR LANE MANAGEMENT

A Presentation on. Human Computer Interaction (HMI) in autonomous vehicles for alerting driver during overtaking and lane changing

Performing a More Realistic and Complete Safety Analysis by Means of the Six-Variable Model. Nelufar Ulfat-Bunyadi, Denis Hatebur, Maritta Heisel

COMPREHENSIVE COMPONENTS MONITORING

Experience the Hybrid Drive

Functional Algorithm for Automated Pedestrian Collision Avoidance System

SIRIUS 2001 A Drive-by-Wire University Project

An Integrated Process for FDIR Design in Aerospace

FUTURE BUMPS IN TRANSITIONING TO ELECTRIC POWERTRAINS

Embedded Torque Estimator for Diesel Engine Control Application

Contents. Preface... xiii Introduction... xv. Chapter 1: The Systems Approach to Control and Instrumentation... 1

MULTIBODY ANALYSIS OF THE M-346 PILOTS INCEPTORS MECHANICAL CIRCUITS INTRODUCTION

NASA Glenn Research Center Intelligent Power System Control Development for Deep Space Exploration

Low Carbon Vehicle Technology Program

Research Challenges for Automated Vehicles

The TIMMO Methodology

Automated Driving - Object Perception at 120 KPH Chris Mansley

Model Based Design: Balancing Embedded Controls Development and System Simulation

Research on Failure mode and effect analysis of Diesel Engine

Incorporating Drivability Metrics into Optimal Energy Management Strategies for Hybrid Vehicles. Daniel Opila

Good Winding Starts the First 5 Seconds Part 2 Drives Clarence Klassen, P.Eng.

Centerwide System Level Procedure

QS 100 LSM Power Management

Using SystemVerilog Assertions in Gate-Level Verification Environments

CMMI Opens the Gate: The Practical Relationship between CMMI and Stage- Gate Decision Models

The Airline Industry Delta Air Lines, Inc. Technical Operations Engine Maintenance Operations

Test & Validation Challenges Facing ADAS and CAV

Session Four Applying functional safety to machine interlock guards

SUBJECT: Automatic Stability Control with Traction Control System (ASC+T)

Optimizing Battery Accuracy for EVs and HEVs

MORSE: MOdel-based Real-time Systems Engineering. Reducing physical testing in the calibration of diagnostic and driveabilty features

Hybrid Architectures for Automated Transmission Systems

SIRIUS A Drive-by-Wire University Project. Per Johannessen Chalmers University of Technology Volvo Car Corporation

Proper Modeling of Integrated Vehicle Systems

FUNCTIONAL SAFETY FOR AUTONOMOUS DRIVING

three different ways, so it is important to be aware of how flow is to be specified

Use of Flow Network Modeling for the Design of an Intricate Cooling Manifold

Low Carbon Technology Project Workstream 8 Vehicle Dynamics and Traction control for Maximum Energy Recovery

Simulated EV Dynamics: Safety & etvc

L (LU4, LJ3, L88) used in Saab 9-5 ENGINE DIAGNOSTIC PARAMETERS

Control as a Service (CaaS)

Software Failure Analysis of Brake-By-Wire Automotive Safety Critical System using FMEA, FTA and MATLAB Techniques

Momentu. Brake-by-Wire Gathers. HIL Test System for Developing a 12-V Brake-by-Wire System BRAKE-BY-WIRE SYSTEMS

GRID MODERNIZATION INITIATIVE PEER REVIEW GMLC Control Theory

Offshore Application of the Flywheel Energy Storage. Final report

Safety Exhaust Valve Integration Guide

Robust Design Workshop Technical University of Denmark Robust Design Day 8 th of November 2017

Autonomous People Mover P15241

Five Cool Things You Can Do With Powertrain Blockset The MathWorks, Inc. 1

UNCLASSIFIED FY 2017 OCO. FY 2017 Base

Model Based Development and Calibration

Energy System Design for Optimized Power Management

ASI-CG 3 Annual Client Conference

HIGH-RELIABILITY POUCH CELL CONNECTION AND COST ASPECTS OF A ROBUST BMS SOLUTION

Table of Contents 1. INTRODUCTION GENERAL INFORMATION-ABOUT OBDII/EOBD PRODUCT DESCRIPTIONS OPERATIONS...11

Problem Definition Review

Cooperative Autonomous Driving and Interaction with Vulnerable Road Users

Test Plans & Test Results

Testing Electrified Drivetrains for Vehicles without the Battery or Engine. Application Reprint of Readout No. 38

Accident Reconstruction & Vehicle Data Recovery Systems and Uses

Siemens PLM Software develops advanced testing methodologies to determine force distribution and visualize body deformation during vehicle handling.

Automatized System of Electrical Diagnosis Validation

Discovery of Design Methodologies. Integration. Multi-disciplinary Design Problems

Design and Calibration of the Jaguar XK Adaptive Cruise Control System. Tim Jagger MathWorks International Automotive Conference 2006

THE ELECTRIC AUTOMOTIVE MOTOR CONTROL SYSTEM

Automotive Electronics/Connectivity/IoT/Smart City Track

POWERTRAIN SOLUTIONS FOR ELECTRIFIED TRUCKS AND BUSES

KISSsys Application 008: Gearbox Concept Analysis

E-DRIVE: HIGHLY INTEGRATED AND HIGH EFFICIENT

NETSSWorks Software: An Extended AC Optimal Power Flow (AC XOPF) For Managing Available System Resources

Qualification Testing of High Rate of Fire Gun Systems

Modification of IPG Driver for Road Robustness Applications

Transcription:

Compatibility of STPA with GM System Safety Engineering Process Padma Sundaram Dave Hartfelder

Table of Contents Introduction GM System Safety Engineering Process Overview Experience with STPA Evaluation procedure STPA Steps 1. Identify Hazards Apply PHA which includes vehicle functional HAZOP analysis 2. Draw Control Structure 3. Identify Unsafe Control Actions, Safety Constraints 4. Causal Factors, refine detailed safety requirements GM Safety process steps to derive safety requirements Summary Results Conclusion

Introduction Electronics and software content continue to increase in automotive systems Safety-critical systems require disciplined and comprehensive engineering effort to identify safety related risks and eliminate or control them Need to address both random and systematic concerns GM has been engaged in engineering safety-critical systems since the early 1980 s Internally developed robust processes have been put in place to verify the integrity of these systems since the launch of electronic throttle control (ETC) in 1997 Safety process influenced by MIL STD 882 The product engineering processes are continuously enhanced to handle growing complexity of systems, optimize efficiency and also to be aligned with the ISO 26262

System Safety- Random and Systematic systematic failure (Per IEC 61508 definition): failure related in a deterministic way to a certain cause, which can only be eliminated by a modification of the design or of the manufacturing process, operational procedures, documentation or other relevant factors Examples of causes of systematic failures include human error in the safety requirements specification; the design, manufacture, installation, operation of the hardware; the design, implementation, etc. of the software. Process Issue System Malfunctions Systematic Causes Design Flaws Incorrect Requirement (s) Missing Requirement (s) Omission of Requirements Example Potential Causes: 1. Incomplete understanding of the system behavior under all operating conditions 2. Engineering Process flaws 3. Management/Communication flaws Random Causes Failure of hardware devices including Sensors, Processors, Actuators, connectors, Communication Interfaces due to mechanical wear, ageing, stress

GM System Safety Engineering Process

STPA Evaluation Study As part of continuous improvement of our process, we are open to evaluating any new technique that can help us improve the effectiveness and efficiency of the safety process It is in this context that we decided to investigate STPA as applied to an embedded automotive control system We have applied STPA on a simplified Engine Control system and compared the derived safety requirements against those derived by following GM system safety process steps Two teams involved in the study One performed the STPA and derived the safety requirements Other followed the GM safety process to derive safety requirements Both teams got together to compare the safety requirements objectively

Source: STPA/STAMP Workshop #1, April 2012, MIT STPA Steps

To identify Hazards, GM PHA process was followed: Motion Control Vehicle Level Hazards (Example) Note: In this presentation, only motion control vehicle hazards are being considered Source: SAE Vehicle Axes

Generic Vehicle Level Hazards Based on the 3 primary degrees of freedom of vehicle, the Table on the right shows the generic vehicle level motion control hazards that are possible Hazards are based on motion control properties that can be potentially affected by malfunctioning electrical/electronic control systems Not all of the hazards are applicable to all vehicle systems To identify the applicable hazards for a given vehicle system, GM s system safety process employs a HAZOP analysis to map malfunction behaviors to vehicle level hazards This is discussed in coming slides Potential Vehicle Level Hazards (Example) Unintended Longitudinal Vehicle Acceleration Loss/Reduced Longitudinal Vehicle Acceleration Unintended Travel in the Opposite Direction Unintended Propulsion Engage (or Power flow) Loss of Propulsion (or Power flow) Unintended Longitudinal Vehicle Motion (Rollaway) Loss of Longitudinal Vehicle Motion Unintended Vehicle Deceleration Loss/Reduced Vehicle Deceleration Unintended Lateral Vehicle Motion Loss of Lateral Vehicle Motion Unintended Vehicle Yaw Unintended Vehicle Vertical Motion/Roll

PHA: To identify system specific hazards Define System Identify System Malfunctions Map Vehicle Behaviors Assess Risk

Simplified Engine Control System Driver Pedal Sensor Sensor Info Engine Controller Actuator Command Actuators Actuator Output ICE Propulsion Torque External Controller(s) Other System Information Feedback sensors via Flywheel, Transmission, Gear Ratio Feedback information

PHA: To identify system specific hazards Define System Identify System Malfunctions Map Vehicle Behaviors Assess Risk HAZOP Approach

Identifying Malfunction Behaviors: HAZOP Methodology Guidewords System Function Vs. Guidewords Loss of Function Incorrect Function -i (More than design intent) Incorrect Function- ii (Less than Design intent) Incorrect Function- iii (Wrong direction) Unintended Activation (Incorrect Timing) Locked/Stuck Function Control Propulsion Torque Loss of Propulsion Torque Excessive Propulsion Torque Reduced Propulsion Torque - Unintended Propulsion Torque Stuck Propulsion (Stuck Throttle)

PHA: To identify system specific hazards Define System Identify System Hazards Map to Vehicle Hazards Assess Risk

Map Malfunction Behaviors to Vehicle Hazards Hazards for Engine Control System (Example) Malfunction Behaviors Unintended Propulsion Torque Excessive Propulsion Torque Vehicle Hazards Unintended Acceleration Stuck Propulsion Torque Loss of Propulsion Torque Reduced Propulsion Torque Unexpected Engine Start when vehicle is static Loss of Acceleration Reduced Acceleration Access to Rotating Components during Service

PHA: To identify system specific hazards Define System Identify System Hazards Map Vehicle Behaviors Assess Risk Determine ASIL

Risk Assessment Per ISO26262, risk is expressed in terms of an Automotive Safety Integrity Level (ASIL) ASIL = function of (S, E, C) S = Severity of the Hazard E = Likelihood of Exposure to the operating scenario C = Controllability of the operator/involved people Unable to avoid the mishap ASIL specifies the developmental process rigor, the required hardware and software integrity requirements for the safety-critical system For each identified hazard, a safety goal, and ASIL is specified. This is translated into one or more vehicle level safety requirements. For example, for a propulsion system, a vehicle level safety requirement could be unintended longitudinal acceleration change at a vehicle level shall not exceed x g in y sec.; Design shall meet ASIL D integrity requirements for process and product

Source: STPA/STAMP Workshop #1, April 2012, MIT STPA Steps

System Control Structure Driver Determine Axle Torque Request; Control Actuators (Throttle, Fuel, Spark, etc.); Pedal Sensor Sensor Info Engine Controller Actuator Command Actuators Actuator Output ICE Propulsion Torque Drive Wheels External Controller(s) Other System Information Feedback sensors via Flywheel, Transmission, Gear Ratio Feedback Information

Source: STPA/STAMP Workshop #1, April 2012, MIT STPA Steps

Control Action Determine Axle Torque Request Action Required but not provided; Not providing causes a hazard Axle Torque request and other relevant inputs not determined; Control system does not know what the current axle torque request is; Possibly uses the previous request or default(unknown) request; Potential for incorrect actuator command leading to unintended or excessive torque output- hazard. Unsafe Action Provided; Providing causes a hazard Axle Torque request calculation incorrect; Control system miscalculates the Axle Torque request; Potential for incorrect throttle command leading to unintended or excessive torque output- hazard Unsafe Control Actions Incorrect Timing Order Periodic Axle Torque Request computed too slow; Will lead to imprecise output from the control system; Output may not match input precisely. Potential for incorrect throttle command leading to unintended or excessive torque outputhazard Stopped Too Soon Axle Torque request determination stopped; If the Axle Torque request determination stops, and the system is not aware, then the effect is similar to pedal request not being calculated; Potential for incorrect actuator command leading to unintended or excessive torque outputhazard. Command Engine Actuators Actuator Control not performed; Control System does not control the actuators to the required position; Actuator may be stuck at a position and not changed; Potential for incorrect throttle command leading to unintended, excessive or stuck torque outputhazard Control system opens the throttle wide when it should be closed; May be Control Throttle is stuck; Potential for incorrect throttle command leading to unintended, excessive or stuck torque output- hazard Actuator control performed too late; May not match driver/axle torque request; Output will lag the driver/axle torque request if the control is delay; Potential for incorrect throttle command leading to unintended, excessive or stuck torque outputhazard Actuator control stopped; Output may be zero when driver request is not; If the control propulsion Output stops, then the output is not controlled. System propulsion output could be reduced to zero. Loss of vehicle acceleration;

Defining Safety Constraints Unsafe Control Actions Axle Torque request and other relevant inputs not determined; High Level Safety Constraints The Control system shall determine the axle torque request periodically Axle Torque request calculation incorrect; The Control System shall not miscalculate the Axle Torque request Axle Torque Request computed too slow; Output Control shall be synchronized with Input request Axle Torque request determination stopped; Actuator Control not performed; Actuator Control performed when it should not; System shall be made aware of the Axle Torque Command determination status System shall be able to control the actuator (Throttle, Spark, Fuel, etc.), when required; Control system shall control the actuator (Throttle, Spark, Fuel, etc.) correctly in response to axle torque request; Actuator control performed too late; Actuator control stopped; Actuator (Throttle, Spark, Fuel, etc.) Control shall be synchronized with axle torque request; If the control propulsion output stops, System propulsion output shall be gradually ramped to zero with operator notification. Graceful transition to a safe state

Source: STPA/STAMP Workshop #1, April 2012, MIT STPA Steps

Driver Incorrect or missing or delayed sensor information EMI Pedal Sensor Sensor Info STPA Analysis: Causal Factors Incorrect/Inadequate control of safety critical outputs Incorrect requirements, design flaws, Calibration changes, hardware integrity issues Determine Axle Torque Request; Control Actuators (Throttle, Fuel, Spark, etc.); Input Processing Input Arbitration Engine Controller Compute Cmnds Actuator Command Actuators Component Faults, changes overtime, unidentified or out-of range disturbances, Stuck Control, mechanical issues, EMI Component faults, changes overtime, mechanical issues, disturbances due to placement in the vehicle External Controller(s) Other System Information Incorrect or missing external information Requirements, design flaws, Changes, hardware integrity issues Control Outputs Incorrect/delayed/ inadequate operation Feedback sensors Actuator Output Feedback Information Incorrect or missing or delayed sensor information Requirements, design flaws, Changes, hardware integrity issues ICE Propulsion Torque via Flywheel, Transmission, Gear Ratio Component Faults, changes overtime, unidentified or out-of range disturbances Drive Wheels

STPA Analysis: Refined Safety Requirements STPA Derived Safety Requirements Control system shall compute periodically the Axle Torque request and other relevant inputs; The Control System shall correctly calculate Axle Torque request based on all of it s inputs Output control timing shall be synchronized with axle torque request; (Output should be in sync. with input) Integrity of relevant feedback information shall be verified; System shall be able to correctly control the actuator (Throttle, Spark, Fuel, etc.) when required; Actuator stuck shall be diagnosed; Control system shall control the actuator (Throttle, Spark, Fuel, etc.) correctly in response to axle torque request; Feedback information shall be diagnosed for integrity; Actuator (Throttle, Spark, Fuel, etc.) Control timing shall be synchronized with Axle Torque Request; (Output should be in sync. with input); Integrity of relevant feedback Information shall be verified to ensure actuator (Throttle, Spark, Fuel, etc.) control output is following axle torque request; Hardware integrity metrics compliance for sensors, actuators, motor control and driver, i/o circuit, feedback information, processor and controller hardware circuit shall be met; Process to check correctness of System, Software and Hardware requirements, design, implementation and V&V; Process to verify integrity checks during software changes, calibration changes; EMI Analysis; Ergonomics Human factors analysis for Pedal placement shall be performed.

Table of Contents Introduction GM System Safety Engineering Process Overview Experience with STPA Evaluation procedure STPA Steps 1. Identify Hazards Apply PHA which includes vehicle functional HAZOP analysis 2. Draw Control Structure 3. Identify Unsafe Control Actions, Safety Constraints 4. Causal Factors, refine detailed safety requirements GM Safety process steps to derive safety requirements steps Summary Results Conclusion

Concept Phase GM Safety Process: Safety Requirements Derivation System Safety Program Plan Directs and manages the safety process execution Requirements Phase Refine requirements Design Phase

Requirement at a vehicle level Requirement at a Control System level Unintended acceleration change shall not exceed x g within t ms ; Hazard Metric Translation for a given vehicle, control system Example Safety Requirements Derivation Incorrect Axle Torque shall not exceed +/-y Nm within z ms ASIL also specifies the process integrity requirements Function 1 Axle Torque Calculation Requirements allocation to software and hardware Function 2 Actuator Command Control HAZOP is applied to functional interfaces to identify safety critical interfaces and interactions Sensor Inputs Integrity Controller Processor Integrity External Controller Input(s) Integrity Communication Integrity Integrity Requirements for Components Control Outputs Integrity Control Feedback Sensors Integrity Actuator Component Integrity Power Supply Integrity Tailored System FTA and Software functional HAZOP analysis (SAE ARP 5580) are used considering the control propagation paths to verify requirement completeness

Table of Contents Introduction GM System Safety Engineering Process Overview Experience with STPA Evaluation procedure STPA Steps 1. Identify Hazards Apply PHA which includes vehicle functional HAZOP analysis 2. Draw Control Structure 3. Identify Unsafe Control Actions, Safety Constraints 4. Causal Factors, refine detailed safety requirements GM Safety process steps to derive safety requirements steps Summary Results Conclusion

Analysis Results Summary The STPA derived requirements were compared to the those derived by the GM system safety process Overall safety requirements derived from the GM safety analysis activities were compatible with the requirements from STPA This may be attributed to the reason that GM system safety process has HAZOP analysis, and tailored FTA approaches that considers the control function in it s analysis at different levels The safety team identified that there could be opportunity for further specificity in the detailed safety requirements STPA derived requirements/constraints in terms of unsafe control actions were readily allocatable Causal factors considers both systemic and random causes. This enables us to specify requirements to address all the causes This opportunity to be further explored with specific recommendations developed by late 2013 Considering similar study using external resources in 2013

Conclusions STPA technique is valuable and different from other techniques such as traditional FTA and FMEA If FTA or FMEA focused only on the physical architecture without consideration to control system propagation paths and feedback mechanisms, it may be possible to miss some safety requirements GM FTA approach is tailored to include control system propagation paths within the hardware and software architecture HAZOP analysis of software architecture considers control propagation path Overall safety requirements derived from GM Concept and Requirements Phase safety analysis activities were compatible with the requirements from STPA GM safety team is continuing to study the benefits of incorporating STPA to enhance its safety engineering process STPA inclusion could help verify completeness of safety requirements earlier in the process Specific recommendations to be developed by late 2013

Acknowledgements Thanks to the support of GM Engine Control Safety team members Rich Kulas and Tim Hartrey for supporting this study

Thank You

Backup

Safety Analysis Techniques Start with the known causes Inductive Reasoning Possible effects Possible causes Deductive Reasoning Start with the known Effects Possible causes Exploratory Reasoning Start with single deviation Exploratory Reasoning Possible effects

Identifying Malfunction Behaviors: HAZOP Approach Hazard Operability (HAZOP) Approach: Use Guidewords to guide the analysis Start with each system function and consider the following system behaviors System function not provided when needed System function provided when not needed System function provided incorrectly when needed Excessive- more than design intent Inadequate- less than design intent Different direction- in the opposite direction System function locked/frozen

System Level SW HAZOP Analysis (SAE ARP 5580) Four basic guidewords for each software element Fails to execute Executes incompletely Functionally erroneous Executes with incorrect timing Too early, too late, takes to long to complete Software interfaces are evaluated Input errors - logically complete set Output errors - logically complete set