近年來，國際間希望藉由感知技術、AI演算法…應用發展自動駕駛，並透過V2X車聯網通訊技術，藉以降低人為疏失之交通事故發生機率。我國以自駕巴士為自駕技術之發展目標，「安全」為測試發展關鍵目標，因此本研究先彙整國際自動駕駛系統安全性規範，並建構自駕系統安全性評估架構，然後藉由修正式德爾菲藉專家知識與經驗確立評估架構。本研究進一步設計自駕巴士行駛安全性風險評估問卷，並以「新北巿淡海智駕電動巴士環線多車服務測試運行計畫」為實證分析對象，建立「變換車道」、「路口左轉彎」、「行經行人穿越道」三項場景，在自駕系統安全性評估架構下探討自駕巴士於不同場景下影響行駛之安全性因素。本研究基於「行駛環境」、「系統安全」、「V2X車聯網安全」、「駕駛員接管狀態」四大構面下16項風險因素，並以ISO 26262規範之車輛安全完整性等級（Automotive Safety Integrity Level ,ASIL）之「嚴重度」、「暴露」、「可控性」作為風險評估指標，借助自駕系統與智慧運輸專家之知識與經驗予以評估而找出影響各場景之關鍵安全風險因素。
本研究先以熵權法進行風險評估指標之權重計算，再以TOPSIS進行各場景風險因素排序，結果得出三場景之中「駕駛員接管狀態」、「行駛環境」為關鍵影響構面，「駕駛員接管狀態」下之關鍵風險因素為「駕駛經驗少」、「HMI傳遞訊息延遲」、「車流密度高」，「行駛環境」中「雨天影響感測」、「光線影響感測」為關鍵因素，因此評估自駕巴士行駛安全時，駕駛培訓與人機介面及環境評估為主要影響因素。

In recent years, autonomous vehicles which can reduce the occurrence of accidents caused by human factors have been developed by using perception technology, AI algorithms and V2X communication technology worldwide. The development of autonomous bus is on the way in Taiwan, however, safety is still the first priority issue that must be discussed. This study firstly overviews international safety standards or regulations of autonomous vehicles. A system safety assessment framework for autonomous vehicles is then proposed and the safety assessment framework is finalized by using the modified Delfi method to reach consensus of experts.
This study further designs a safety risk assessment questionnaire for autonomous buses. The case of "Pilot Test of New Taipei City Tamhai Smart Bus for Loop Multi-Vehicle Service" is chosen for empirical studies.  This study explores the possible safety factors affecting autonomous buses in three scenarios including "changing lanes", "turning left at intersections", and "Pedestrian Crossing" under the safety assessment framework. Based on "driving environment", "system security", "V2X Internet of Vehicles security" and  "Driver Takeover Status" four dimensions and 16 risk factors,  the "severity", "exposure" and "controllability" of the vehicle safety integrity level (Automotive Safety Integrity Level, ASIL) of the ISO 26262 standard are used as risk assessment indicators. Then autonomous vehicle and smart transportation experts are invited to evaluate and identify key safety risk factors affecting each scenario.
The entropy weight method is used to calculate the weight of risk assessment indicators, and then TOPSIS methos is used to rank the risk factors of each scenario. The results show that among the three scenarios, "driver takeover status" and "driving environment" are the key influencing dimensions. The key risk factors in the "driver takeover status" dimension are "few driving experience", "HMI message transmission delay" and "high traffic density". In the "driving environment" dimension, "rainy weather affects sensing" and "light affects sensing" are key influencing factors. It is concluded that driver training, human-machine interface and environment factors will mainly influence the safety of autonomous buses.

目錄	I
圖目錄	IV
表目錄	V
第一章 緒論	1
1.1 研究背景與研究動機	1
1.2 研究目的	2
1.3 研究範圍	3
1.4 研究流程	4
第二章 文獻回顧	5
2.1 自動駕駛定義	5
2.2 車聯網發展現況	6
2.3 自駕系統安全性規範	7
2.4 自駕系統運行安全性相關研究	14
2.4.1 操作適用範圍(ODD)定義	14
2.4.2 系統安全與接管措施	19
2.4.3 V2X車聯網安全	22
2.4.4 法律安全性規範	24
2.5 我國交通部自駕公車實驗安全指引之推動政策	26
2.6 自駕巴士行駛安全風險因素探討	28
2.6.1 行駛環境	29
2.6.2 系統安全	32
2.6.3 V2X車聯網安全	33
2.6.4 駕駛員接管狀態	34
2.7 修正式德爾菲法	35
2.7.1 修正式德爾菲法應用	37
2.8 TOPSIS熵權法	38
2.9 自動駕駛測試評估方法	41
2.10 場景產生方法	43
2.11 文獻回顧小結	46
第三章 研究方法	48
3.1 研究架構	48
3.2 整體系統評估構面與準則擬定	49
3.3 研究方法	51
3.3.1 修正式德爾菲法	51
3.3.2 TOPSIS熵權法	52
3.4 自駕巴士運行安全性風險評估問卷設計	55
3.4.1 評估場景擬定與描述	55
3.4.2 測試車輛基本資料	61
3.4.3 自駕巴士行駛安全風險因素探討	62
3.4.4 自駕巴士行駛安全風險評估指標	72
第四章 實證分析	76
4.1 修正式德爾菲法	76
4.2 TOPSIS熵權法風險評估	77
4.2.1 路口左轉場景	82
4.2.2 行人穿越道場景	84
4.2.3 變換車道場景	85
4.2.4 專家群觀點之異同分析	86
4.2.5 管理意涵	90
第五章 結論與建議	94
5.1 結論	94
5.2 建議	95
參考文獻	96
附錄一 自駕巴士運行安全性評估修正式德爾菲問卷	101
附錄二 自駕巴士運行安全性風險評估問卷	107


圖目錄

圖 1.1 淡海第二期計畫路線圖	3
圖 1.2 研究流程圖	4
圖 2.1 NHTSA ODD 規範架構	15
圖 2.2 BSI ODD 規範架構	15
圖 2.3 SAE AVSC ODD 規範架構	15
圖 2.4 場景與功能測試整合評估架構	42
圖 2.5 情境與情景文氏圖(Venn diagram)	44
圖 2.6 情境與情景案例說明	45
圖 3.1 研究架構圖	48
圖 3.2 整體系統評估架構	50
圖 3.3 路口左轉場景示意圖	59
圖 3.4 行人穿越道場景示意圖	60
圖 3.5 變換車道場景示意圖	60
圖 3.6 車輛感測器配置圖	62
圖 3.7 自駕巴士安全性評估架構與安全性風險因素之關係圖	65
圖 4.1 自駕巴士安全性之風險因素評估架構圖	78


表目錄
表 2.1 SAE J3016 車輛自動化級別	6
表 2.2 車聯網應用類型	7
表 2.3 日本「自動駕駛汽車安全技術指南」內容彙整	8
表 2.4 德國「德國自動化與聯網化駕駛策略」內容彙整	10
表 2.5 UN ECE R157 ALKS 規範內容彙整	11
表 2.6 國際自動駕駛之安全性規範彙整	13
表 2.7 操作適用範圍(ODD)定義類型	18
表 2.8 OEDR 行為類型綜整	20
表 2.9「自駕公車實驗運行安全指引」內容彙整	27
表 2.10「自駕公車實驗運行安全指引」對應國際自駕安全規範分類彙整	28
表 2.11 自駕巴士行駛安全風險因素探討內容	29
表 3.1 台北市 110 年交通事故前十大肇因統計	55
表 3.2 自駕巴士運行安全性風險評估場景彙整	56
表 3.3 各場景六層模型基本資料綜整	57
表 3.4 測試車輛基本規格表	61
表 3.5 感測系統元件彙整表	61
表 3.6 行駛環境之風險因素定義與可能危害	66
表 3.7 系統安全之風險因素定義與可能危害	68
表 3.8 各系統故障機率	69
表 3.9 V2X 車聯網安全之風險因素定義與可能危害	69
表 3.10 駕駛員接管狀態之風險因素定義與可能危害	71
表 3.11 ISO 26262 標準嚴重度等級描述	72
表 3.12 ISO 26262 標準暴露等級描述	73
表 3.13 ISO 26262 標準可控性等級描述	74
表 3.14 Da?deviren, M., et al. (2009) TOPSIS 評估尺度	74
表 4.1 修正式德爾菲問卷專家名單	76
表 4.2 自駕系統整體安全評估準則篩選結果	76
表 4.3 TOPSIS 熵權法風險評估問卷專家名單	77
表 4.4 各場景決策矩陣建立結果	78
表 4.5 各場景標準化矩陣建立結果	79
表 4.6 各場景計算指標權重結果	80
表 4.7 各場景加權歐氏距離?????? +計算結果	81
表 4.8 各場景加權歐氏距離?????? ?計算結果	81
表 4.9 各場景風險因素相對接近程度????計算結果	82
表 4.10 路口左轉場景風險評估指標計算結果	83
表 4.11 路口左轉場景風險因素排序結果	83
表 4.12 行人穿越道場景風險評估指標計算結果	84
表 4.13 行人穿越道場景風險因素排序結果	84
表 4.14 變換車道場景風險評估指標計算結果	85
表 4.15 變換車道場景風險因素排序結果	85
表 4.16 路口左轉場景專家群獨立分析結果	86
表 4.17 行人穿越道場景專家群獨立分析結果	87
表 4.18 變換車道場景專家群獨立分析結果	89

1. Automated Vehicle Safety Consortium. (2020). AVSC Best Practice for 
Describing an Operational Design Domain: Conceptual Framework and 
Lexicon. SAE Industry Technologies Consortia.
2. Amditis, A., Lytrivis, P., & Manganiaris, S. (2020). Infrastructure Supported 
Operational Design Domain: ISAD closing ODD gaps. In Virtual ITS European 
Congress.
3. Adee, A., Gansch, R., & Liggesmeyer, P. (2021, September). Systematic 
modeling approach for environmental perception limitations in automated 
driving. In 2021 17th European Dependable Computing Conference 
(EDCC) (pp. 103-110). IEEE.
4. Brorsson, A. (2022). Challenges Within V2X: A cybersecurity risk assessment 
for V2X use cases.
5. Cafiso, S., Di Graziano, A., & Pappalardo, G. (2013). Using the Delphi method 
to evaluate opinions of public transport managers on bus safety. Safety 
science, 57, 254-263.
6. Cui, J., Sabaliauskaite, G., Liew, L. S., Zhou, F., & Zhang, B. (2019). 
Collaborative analysis framework of safety and security for autonomous 
vehicles. IEEE Access, 7, 148672-148683.
7. Chen, F., Lu, G., Lin, Q., Zhai, J., & Tan, H. (2021). Are novice drivers 
competent to take over control from level 3 automated vehicles? A comparative 
study with experienced drivers. Transportation research part F: traffic 
psychology and behaviour, 81, 65-81.
8. Chen, C., Qidong, Z., Tong, Z., Yang, Z., & Xianglei, Z. （2022）. The 
Research on Current Automated Driving ODD Regulations, Standards and 
Applications. In 2022 IEEE International Conference on Real-time Computing 
and Robotics （RCAR） （pp. 744-747）. IEEE.
9. Das, P. (2018). Risk analysis of autonomous vehicle and its safety impact on 
mixed traffic stream. Rowan University.
10. den Hartog, J., & Zannone, N. (2018). Security and privacy for innovative 
automotive applications: A survey. Computer Communications, 132, 17-41.
11. Dehdasht, G., Ferwati, M. S., Zin, R. M., & Abidin, N. Z. (2020). A hybrid 
approach using entropy and TOPSIS to select key drivers for a successful and 
sustainable lean construction implementation. PloS one, 15(2), e0228746.
12. Formosa, N., Quddus, M., Man, C. K., Singh, M. K., Morton, C., & Masera, C. 
B. (2023). Evaluating the Impact of Lane Marking Quality on the Operation of 
Autonomous Vehicles. Journal of Transportation Engineering, Part A: 
Systems, 150(1), 04023126.
13. Gosavi, M. A., Rhoades, B. B., & Conrad, J. M. (2018). Application of 
functional safety in autonomous vehicles using iso 26262 standard: A survey. 
In SoutheastCon 2018 (pp. 1-6). IEEE.
14. Guo, J., Kurup, U., & Shah, M. (2019). Is it safe to drive? an overview of 
factors, metrics, and datasets for driveability assessment in autonomous 
driving. IEEE Transactions on Intelligent Transportation Systems, 21(8), 3135-
3151.
15. Ghosal, A., & Conti, M. (2020). Security issues and challenges in V2X: A 
survey. Computer Networks, 169, 107093
16. Huang, J. (2008). Combining entropy weight and TOPSIS method for 
information system selection. In 2008 ieee conference on cybernetics and 
intelligent systems (pp. 1281-1284). IEEE
17. Hobert, L., Festag, A., Llatser, I., Altomare, L., Visintainer, F., & Kovacs, A. 
(2015). Enhancements of V2X communication in support of cooperative 
autonomous driving. IEEE communications magazine, 53(12), 64-70.
18. Heinzler, R., Schindler, P., Seekircher, J., Ritter, W., & Stork, W. (2019). 
Weather influence and classification with automotive lidar sensors. In 2019 
IEEE intelligent vehicles symposium (IV) (pp. 1527-1534). IEEE.
19. Hendrik Weber, Julian Bock, Jens Klimke, Christian Roesener, Johannes Hiller, 
Robert Krajewski, Adrian Zlocki & Lutz Eckstein （2019） A framework for 
definition of logical scenarios for safety assurance of automated driving, Traffic 
Injury Prevention, 20:sup1, S65-S70, DOI: 10.1080/15389588.2019.1630827
20. Ito, M. (2021). ODD description methods for automated driving vehicle and 
verifiability for safety. J. Univers. Comput. Sci., 27(8), 796-810.
21. Jozi, S. A., Shafiee, M., MoradiMajd, N., & Saffarian, S. (2012). An integrated 
Shannon's Entropy–TOPSIS methodology for environmental risk assessment of 
Helleh protected area in Iran. Environmental monitoring and assessment, 184, 
6913-6922.
22. Kuehn, M., Vogelpohl, T., & Vollrath, M. (2017, May). Takeover times in 
highly automated driving (level 3). In 25th International technical conference 
on the enhanced safety of vehicles (ESV) national highway traffic safety 
administration (pp. 1-11).
23. Koopman, P., & Fratrik, F. (2019). How many operational design domains, 
objects, and events?. Safeai@ aaai, 4.
24. Kawser, M. T., Fahad, M. S., Ahmed, S., Sajjad, S. S., & Rafi, H. A. (2019). 
The perspective of vehicle-to-everything (v2x) communication towards 
5g. IJCSNS, 19(4), 146.
25. Koulinas, G. K., Demesouka, O. E., Sidas, K. A., & Koulouriotis, D. E. (2021). 
A TOPSIS—risk matrix and Monte Carlo expert system for risk assessment in 
engineering projects. Sustainability, 13(20), 11277.
26. Li, L., Huang, W. L., Liu, Y., Zheng, N. N., & Wang, F. Y. （2016）. 
Intelligence testing for autonomous vehicles: A new approach. IEEE 
Transactions on Intelligent Vehicles, 1（2）, 158-166.
27. Liu, Y., Tight, M., Sun, Q., & Kang, R. (2019). A systematic review: Road 
infrastructure requirement for Connected and Autonomous Vehicles (CAVs). 
In Journal of Physics: Conference Series (Vol. 1187, No. 4, p. 042073). IOP 
Publishing.
28. Linchuan, Y., Yuanyuan, G., & Jixiang, L. (2021). Using Entropy Weight 
Method to Assess Transit Satisfaction: A Focus on Older Adults. China City 
Planning Review, 30(1), 64-73.
29. Marojevic, V. (2018). C-V2X security requirements and procedures: Survey 
and research directions. arXiv preprint arXiv:1807.09338.
30. Mariani, R. (2018). An overview of autonomous vehicles safety. In 2018 IEEE 
International Reliability Physics Symposium (IRPS) (pp. 6A-1). IEEE.
31. Menzel, T., Bagschik, G., & Maurer, M. （2018）. Scenarios for development, 
test and validation of automated vehicles. In 2018 IEEE Intelligent Vehicles 
Symposium （IV） （pp. 1821-1827）. IEEE.
32. Morales-Alvarez, W., Sipele, O., Léberon, R., Tadjine, H. H., & Olaverri-
Monreal, C. (2020). Automated driving: A literature review of the take over 
request in conditional automation. Electronics, 9(12), 2087.
33. National Highway Traffic Safety Administration （NHTSA）. Automated 
Driving Systems （ADS）: A Vision for Safety 2.0. National Highway Traffic 
Safety Administration, U.S. Department of Transportation.
34. No, U. R. (2023). 157—Automated Lane Keeping Systems (ALKS). Nations 
Economic Commission for Europe: Geneva, Switzerland.
35. Ogawa, A., Kuroda, S., Ushida, K., Kudo, R., Tateishi, K., Yamashita, H., & 
Kantou, T. (2018, August). Field experiments on sensor data transmission for 
5g-based vehicle-infrastructure cooperation. In 2018 IEEE 88th Vehicular 
Technology Conference (VTC-Fall) (pp. 1-5). IEEE.
36. Pasagadugula, S., Verma, G., & Harmalkar, J. (2019). Effective approach for 
Redundancy in compliance with ISO26262. In 2019 International Conference on 
Advances in Computing and Communication Engineering (ICACCE) (pp. 1-4). IEEE.
37. Poli, F., Denis, B., Mannoni, V., Berg, V., Martín-Sacristán, D., García-Roger, 
D., & Monserrat, J. F. (2021, April). Evaluation of C-V2X sidelink for 
cooperative lane merging in a cross-border highway scenario. In 2021 IEEE 
93rd Vehicular Technology Conference (VTC2021-Spring) (pp. 1-5). IEEE.
38. PAS, B. (2020). operational design domain (ODD) taxonomy for an automated 
driving system (ADS). Specification. The British Standards Institution, 
Standard.
39. Rau, P., & Thorn, E.（2018） A Framework for Automated Driving System 
Testable Cases and Scenarios.
40. Ren, L., Yin, H., Ge, W., & Meng, Q. (2019). Environment influences on 
uncertainty of object detection for automated driving systems. In 2019 12th 
International Congress on 
41. Saberi, A. K., Luo, Y., Cichosz, F. P., van den Brand, M., & Jansen, S. (2015, 
April). An approach for functional safety improvement of an existing 
automotive system. In 2015 Annual IEEE Systems Conference (SysCon) 
Proceedings (pp. 277-282). IEEE.
42. SAE International. （2016） Surface Vehicle Recommended Practice: （R）
Taxonomy and Definitions for Terms Related to Driving Automation Systems 
for On-Road Motor Vehicles. J3016, Issued January 2014, Revised September 
2016
43. Schildbach, G. （2018）. On the application of ISO 26262 in control design 
for automated vehicles. arXiv preprint arXiv:1804.04349.
44. Sharma, V., You, I., & Guizani, N. (2020). Security of 5G-V2X: Technologies, 
standardization, and research directions. IEEE Network, 34(5), 306-314.
45. Thorn, E., Kimmel, S. C., Chaka, M., & Hamilton, B. A. （2018）. A 
framework for automated driving system testable cases and scenarios （No. 
DOT HS 812 623）. United States. Department of Transportation. National 
Highway Traffic Safety Administration. 
46. Tengilimoglu, O., Carsten, O., & Wadud, Z. (2023). Implications of automated 
vehicles for physical road environment: A comprehensive 
review. Transportation research part E: logistics and transportation review, 169, 
102989.
47. Ulbrich, S., Menzel, T., Reschka, A., Schuldt, F., & Maurer, M. (2015). 
Defining and substantiating the terms scene, situation, and scenario for 
automated driving. In 2015 IEEE 18th international conference on intelligent
transportation systems (pp. 982-988). IEEE.
48. Ü NVAN, Y. A. (2020). Financial performance analysis of banks with TOPSIS 
and fuzzy TOPSIS approaches. Gazi University Journal of Science, 33(4), 904-
923
49. Vreeswijk, J., Wijbenga, A., & Schindler, J. (2020). Cooperative Automated 
Driving for managing Transition Areas and the Operational Design Domain 
(ODD). Proceedings of 8th Transport Research Arena TRA 2020.
50. Vermesan, O., Bahr, R., John, R., Ottella, M., Gjølstad, R., Buckholm, O., & 
Sand, H. E. (2021). Advancing the Design of Fail-Operational Architectures, 
Communication Modules, Electronic Components, and Systems for Future 
Autonomous/Automated Vehicles. In Intelligent System Solutions for Auto 
Mobility and Beyond: Advanced Microsystems for Automotive Applications 
2020 (pp. 53-71). Springer International Publishing.
51. Woori Ko, Sangmin Park, Jaewoong Yun, Sungho Park, & Ilsoo Yun. 
（2022）. Development of a Framework for Generating Driving Safety 
Assessment Scenarios for Automated Vehicles. Sensors （Basel, 
Switzerland）, 22（6031）, 6031–. https://doi.org/10.3390/s22166031
52. Yu, W., Li, J., Peng, L. M., Xiong, X., Yang, K., & Wang, H. (2022). SOTIF 
risk mitigation based on unified ODD monitoring for autonomous 
vehicles. Journal of Intelligent and Connected Vehicles,5(3), 157-166.
53. 日本國土交通省 (2018) ，自動駕駛汽車安全技術指南
54. 李綱. （2019）. 國際車輛自動駕駛技術發展. 土木水利, 46（2）, 11-17.
55. 杜双玉, & 洪念民. (2019). 應用修正式德爾菲法及層級分析法探討臺灣旅
館選址因素. 觀光休閒學報, 25(3), 275-300.
56. 林大傑 、劉欣憲 、周艾蓁 、李佳容 、黃品誠 、吳政欣 、陳銘旭 、張俊
毅（2020）。應用層級分析法建立自駕車測試路線難度評估機制，中華民
國自動機工程學會第二十五屆車輛工程學術研討會論文。
57. 莊靜宜（2021）。我國道路交通法規因應自動駕駛車輛規範法制之研究。
中央警察大學警察政策研究所碩士論文。
58. 財團法人車輛安全審驗中心，自駕公車實驗運行安全指引。
59. 張維容, & 鍾信成. (2009). 以德菲法建立醫院內部評估社區健康服務成效
之指標. 澄清醫護管理雜誌, 5(1), 45-54. 
60. 黃卉庭. (2011). 台灣企業綠色環境績效分析及評估模式之研究-以零售業為
例.
61. 黃哲勳. （2020）. 自駕巴士應用於都市公共運輸的發展契機. 電腦與通訊, 
（183）, 27-29.
62. 馮輝昇, 林俊甫, 邱詩純, & 陳昱辰. （2019）. 水湳智慧城自駕巴士試營運
規劃及展望. 土木水利, 46（2）, 33-38.
63. 楊湘筑, 戴劍鋒, & 彭克仲. (2011). 應用 Entropy 權重法與 TOPSIS 於臺灣
上市食品公司財務績效評估模式之研究. 台灣農學會報, 12(3), 232-249.
64. 劉法旺、徐曉慶、陳貞、等人（2022）搭載自動駕駛功能的智慧網聯汽車
安全測試與評估方法研究。汽車工程學報，12（3）：221 227．
65. 德國交通與數位基礎設施部(2017) ，德國自動化與聯網化駕駛策略執行進
度報告
66. 鄧振源. (2012). 多準則決策分析 : 方法與應用 (一版). 鼎茂圖書出版公司.
67. 陳文亮, & 陳姿樺. (2011). 應用修正式德菲層級程序法建構成衣設計指標
之研究. Journal of Humanities and Social Sciences, 7(1), 49-59.
68. 蔡明田, & 黃昭陽. (2019). 應用層級分析法建構鐵路號誌設備維修策略之
評選模式. 運輸學刊, 31(2), 153-177.
69. 台北市政府資訊局。臺北市信義路公車專用道自駕巴士創新實驗計畫。
70. 網址：https://smartcity.taipei/projdetail/147
71. 新北市政府交通局-新北智駕電動巴士專區
72. 網址：https://www.traffic.ntpc.gov.tw/home.jsp?id=acb536e8e0fbb8ab