SAE J 3018-2015 Guidelines for Safe On-Road Testing of SAE Level 3 4 and 5 Prototype Automated Driving Systems (ADS).pdf

1、 _ SAE Technical Standards Board Rules provide that: “This report is published by SAE to advance the state of technical and engineering sciences. The use of this report is entirely voluntary, and its applicability and suitability for any particular use, including any patent infringement arising ther

2、efrom, is the sole responsibility of the user.” SAE reviews each technical report at least every five years at which time it may be revised, reaffirmed, stabilized, or cancelled. SAE invites your written comments and suggestions. Copyright 2015 SAE International All rights reserved. No part of this

3、publication may be reproduced, stored in a retrieval system or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of SAE. TO PLACE A DOCUMENT ORDER: Tel: 877-606-7323 (inside USA and Canada) Tel: +1 724-776-49

4、70 (outside USA) Fax: 724-776-0790 Email: CustomerServicesae.org SAE WEB ADDRESS: http:/www.sae.org SAE values your input. To provide feedback on this Technical Report, please visit http:/www.sae.org/technical/standards/J3018_201503 SURFACE VEHICLE INFORMATION REPORT J3018 MAR2015 Issued 2015-03 Gui

5、delines for Safe On-Road Testing of SAE Level 3, 4, and 5 Prototype Automated Driving Systems (ADS) RATIONALE This document provides general safety-relevant guidelines for performing tests of prototype automated driving systems (ADSs) equipped on test vehicles operated in mixed-traffic environments

6、on public roads. The levels of automation addressed in this document include conditional (level 3), high (level 4), and full (level 5) as defined by SAE J3016; when activated, these ADSs do not rely on a human driver for monitoring and responding to the vehicle or traffic environment. (SAE J3016 def

7、ines the italicized terms in this document.) These guidelines apply to the testing of all types of motor vehicles including light-duty, passenger, freight or transit vehicles, but are not concerned with component-level testing. These guidelines address only the safety-related prerequisites for on-ro

8、ad testing and conduct of such tests. They do NOT establish performance criteria or test procedures for production vehicles equipped with ADSs. Mixed traffic environments are active public roadways, closed campuses (such as military bases, factories, ports, and enclosed communities), and other conte

9、xts that involve a risk of injury to road users or damage to their property. These environments will (variously) include motor vehicles of all types and classes, pedestrians, and pedal cyclists, as well as animal and object hazards. Accordingly, these guidelines address analytic, laboratory, simulat

10、ion, or closed-course test methods only to the extent that these methods are part of, or predicates to, tests of vehicles equipped with prototype ADSs conducted in mixed-traffic environments on public roadways. The guidelines assume that any such tests will be conducted under some form of human supe

11、rvision; this assumption applies even if the ADS being tested is considered to be a level 5 “full automation” system. Licensing and registration requirements for test vehicles equipped with prototype ADSs should be checked with relevant local jurisdictions. This document also provides guidance on te

12、st driver training and test program management; graduated road testing; test data capture; safety override guidelines, and software development and release requirements. Safety guidelines for the on-road testing of vehicles equipped with prototype ADSs do not currently exist. The information given i

13、n this report will assist interested parties in the creation of guidelines for the on-road testing of such prototypes. SAE INTERNATIONAL J3018 Issued MAR2015 Page 2 of 12 TABLE OF CONTENTS 1. SCOPE 2 1.1 Purpose 2 2. REFERENCES 3 2.1 Applicable Documents . 3 3. DEFINITIONS 3 4. SAFETY TESTING GUIDAN

14、CE . 5 4.1 Test Driver Training 5 4.2 Test Driver Workload . 7 4.3 Managing Test Drivers . 7 4.4 Safety Development Process . 8 4.5 Software Development and Modifications 8 4.6 Selection of Test Routes 8 4.7 Graduated Road Testing 10 4.8 Test Data Capture 10 4.9 Safety Override 11 5. NOTES . 12 5.1

15、Marginal Indicia 12 1. SCOPE This document provides guidelines for the safe conduct of on-road tests of vehicles equipped with prototype conditional, high, and full (levels 3-5) automated driving systems (ADSs), as defined by SAE J3016. It does not include guidance for testing production ADSs intend

16、ed for sale to the general public. The scope is further limited to testing of automated prototype vehicles on public roads. These guidelines do not address: Testing of driver assist (Level 1) or partial (Level 2) automation systems, which rely on a human driver to monitor the environment. (See SAE J

17、3016 for Definitions of Levels of Automated Systems.) Closed-course testing. Component-level testing. The precise regime of road testing for a particular prototype will depend on the intended level of automation and the targeted capabilities of the prototype (see SAE J3016 for more information). A p

18、rototype suitable for testing on public roadways is presumed to have already passed laboratory and/or closed-course testing, which are not addressed by this document. 1.1 Purpose The purpose of this document is to provide information and guidelines for use by automotive designers, test engineers and

19、 policymakers concerned with the safe on-road testing of vehicles equipped with Level 3 (Conditional), Level 4 (High), or Level 5 (Full) ADSs, which are characterized by their ability to perform the complete dynamic driving task whether or not such performance is limited to specific driving modes an

20、d/or roadways/locations. SAE INTERNATIONAL J3018 Issued MAR2015 Page 3 of 12 2. REFERENCES 2.1 Applicable Documents The following publications form a part of this specification to the extent specified herein. Unless otherwise indicated, the latest issue of SAE publications shall apply. 2.1.1 SAE Int

21、ernational Available from SAE International, 400 Commonwealth Drive, Warrendale, PA 15096-0001, Tel: 877-606-7323 (inside USA and Canada) or +1 724-776-4970 (outside USA), www.sae.org. SAE J3016 Taxonomy and Definitions for Terms Related to On-Road Motor Vehicle Automated Driving Systems SAE J1698 E

22、vent Data Recorder SAE J1698-1 Event Data Recorder - Output Data Definition Sheridan, T.B. (1970). Big Brother As Driver: New Demands and Problems for the Man at the Wheel. Human Factors, 12(1), 95-101. 3. DEFINITIONS 3.1 AUTOMATED DRIVING SYSTEM (ADS) (SAE J3016) The hardware and software that is c

23、ollectively capable of performing all aspects of the dynamic driving task for a vehicle (whether part time or full time). 3.2 DRIVING MODE (SAE J3016) A type of driving scenario with characteristic dynamic driving task requirements (e.g., expressway merging, high speed cruising, low speed traffic ja

24、m, etc.). 3.3 DYNAMIC DRIVING TASK (SAE J3016) All of the real-time functions required to operate a vehicle in on-road traffic, excluding the selection of destinations and waypoints (i.e., navigation or route planning) and including without limitation: Object and event detection, recognition, and cl

25、assification; Object and event response; Maneuver planning; Steering, turning, lane keeping, and lane changing; Acceleration and deceleration; Enhancing conspicuity (lighting, signaling and gesturing, etc.). 3.4 EXPERT TEST DRIVER Expert test drivers are specifically skilled and trained personnel wh

26、o are able to activate, monitor, and de-activate the ADS using experimental software and/or hardware interfaces, and who are capable of safely handling the vehicle in a variety of hazardous driving situations. SAE INTERNATIONAL J3018 Issued MAR2015 Page 4 of 12 3.5 HUMAN DRIVER (SAE J3016) The perso

27、n who drives a particular vehicle, and who, in a vehicle equipped with an automated driving system, exchanges the dynamic driving task with such a system as necessary during vehicle operation. 3.6 LEVEL 3 - CONDITIONAL AUTOMATION (SAE J3016) The part-time or driving mode-dependent performance by an

28、automated driving system of all aspects of the dynamic driving task with the expectation that the human driver will respond appropriately to a request to intervene. NOTES: Conditional automation requires a human driver to initiate an automated driving system at the appropriate point during a trip an

29、d requires a human driver to resume the dynamic driving task when s/he receives a request to intervene (e.g., due to a vehicle or system malfunction or a change in driving conditions that exceed the automated driving systems driving mode-dependent capability). As a technical matter, a human driver n

30、eed not monitor the automated driving systems performance while it is engaged, but must be prepared to resume the dynamic driving task when the automated driving system issues a request to intervene. A conditional automated driving system will alert (i.e., by issuing a request to intervene) the huma

31、n driver of the need to resume the dynamic driving task with sufficient time for a typical human driver to respond appropriately. An “appropriate” response by a human driver to a request to intervene may vary depending upon immediate circumstances (e.g., steering, braking, or simply maintaining curr

32、ent input levels), but otherwise entails the timely, safe and correct performance of the dynamic driving task for the prevailing circumstances. 3.7 LEVEL 4 - HIGH AUTOMATION (SAE J3016) The part-time, driving mode-dependent, or geographically-restricted performance by an automated driving system of

33、all aspects of the dynamic driving task, even if a human driver fails to respond appropriately to a request to intervene. NOTES: High automation generally requires a human driver to engage an automated driving system at the appropriate point during a trip (e.g., a human driver may activate the autom

34、ated driving system during a specific driving mode, such as freeway driving, for which the high automated driving system is designed). Examples of exceptions to the general case: - A high automation parking application for which a human driver is not present in the vehicle during the maneuver. - A h

35、igh automation shuttle system that is geographically restricted to operation on a closed or semi-closed campus (e.g., residential community, military base, etc.) As a technical matter, a human driver need not monitor the automated driving systems performance while it is engaged, but should in the ge

36、neral case be prepared to assume performance of the dynamic driving task when the automated driving system issues a request to intervene. A high automated driving system will alert a human driver several seconds in advance of the need to resume the dynamic driving task (i.e., by issuing a request to

37、 intervene); however, the automated driving system is capable of restoring the vehicle to a minimal risk condition automatically if a human driver fails to resume the dynamic driving task when prompted. This capability to automatically restore the vehicle to a minimal risk condition is the only diff

38、erence between high automation and conditional automation, above. SAE INTERNATIONAL J3018 Issued MAR2015 Page 5 of 12 3.8 LEVEL 5 - FULL AUTOMATION (SAE J3016) The unconditional, full-time performance by an automated driving system of all aspects of the dynamic driving task under, at minimum, all ro

39、adway and environmental conditions that can be managed by a human driver, including the ability to automatically bring the motor vehicle into a minimal risk condition in the event of a critical vehicle or system failure, or other emergency event. NOTES: As a technical matter, a human driver need not

40、 monitor the automated driving systems performance. When the automated driving system reaches the limits of its functional capabilities, it will restore the vehicle to a minimal risk condition automatically. 3.9 MONITOR (J3016) The activities and/or automated routines that accomplish comprehensive o

41、bject and event detection, recognition, classification, and response preparation, as needed to competently perform the dynamic driving task. NOTE: When driving vehicles that are not equipped with automated driving systems, human drivers visually sample the road scene sufficiently to competently perf

42、orm the dynamic driving task, while also performing secondary tasks that require short periods of eyes-off-road time (e.g., adjusting cabin comfort settings, scanning road signs, tuning a radio, etc.). Thus, monitoring does not entail constant eyes-on-road time by the human driver. 3.10 NOVICE TEST

43、DRIVER Novice test drivers are licensed, but are not familiar with the ADS and are not trained to activate, monitor, and de-activate it using experimental software and/or hardware interfaces, nor are they trained to safely handle the vehicle in a variety of hazardous driving situations 3.11 SAFE/SAF

44、ETY The absence of unreasonable risk to humans, animals, and property. 3.12 TRAINED TEST DRIVER Trained test drivers are informed about the ADS and are trained to safely handle the vehicle in a variety of hazardous driving situations, but are NOT specifically qualified to activate, monitor, and de-a

45、ctivate the ADS using experimental software and/or hardware interfaces 4. SAFETY TESTING GUIDANCE 4.1 Test Driver Training Driver training is a primary component of a safe testing program because it prepares the test driver for both complex and routine situations likely to be experienced during test

46、-driving. Training ensures that a test driver will know the best actions to take in any situation to minimize risk. In order to react correctly to various situations, a test driver needs to be very familiar with the ADS under test, any applicable malfunction warning systems, software, hardware, and

47、how the vehicle and ADS perform under normal conditions. Initially, test drivers should become familiar with the ADS-equipped vehicle performance characteristics. For example, test driver can participate in informational briefings or classes. After classroom training, hands-on training can be provid

48、ed to prepare the test driver for operating the vehicles under a variety of on-road scenarios. Such hands-on training may introduce drivers to routine maneuvers, road environments, and general operating conditions, and gradually introduce progressively more difficult test procedures, such as emergen

49、cy maneuvers, as test drivers become more experienced. SAE INTERNATIONAL J3018 Issued MAR2015 Page 6 of 12 All drivers are trained to maneuver the general position and speed of the vehicle safely, systematically and smoothly, using road and traffic conditions to progress unobtrusively and skillfully through on-road traffic under a wide range of road and weather conditions. In addition to these rou