Generating and Characterizing Scenarios for Safety Testing of Autonomous Vehicles

Ghodsi, Zahra; Hari, Siva Kumar Sastry; Frosio, Iuri; Tsai, Timothy; Troccoli, Alejandro; Keckler, Stephen W.; Garg, Siddharth; Anandkumar, Anima

doi:10.1109/iv48863.2021.9576023

Cited by 23 publications

(10 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Table 4 provides a comparison of all the simulators used. Note that we omitted NVIDIA DRIVE SIM [89] used in [71] since it is currently in beta version and only provides minimal access and documentation. We next compare these simulators from different perspectives in the following.…”

Section: High-fidelity Simulatormentioning

confidence: 99%

“…Parameters of control for activated behavior control influence an object's activation condition and behavior after being activated. Typical parameters grouped by the type of objects are: (i) NPC vehicles: velocity after certain simulation time [71]; trigger distance for activation [62], [68]; trigger distance for behavior change [65]; velocity after triggered [62], [65]; velocity [57], [67]; and lane change [57], [67] at each time interval.…”

Section: Control For Activated Behaviormentioning

confidence: 99%

“…The min TTC measures the minimal time required for a vehicle to hit an object if both continue with the same speed and do not change their paths. Several more fine-grained measures are proposed in [71] which measure the effort of the ego car to escape from a potential collision by measuring the area of the ego car's potential escape routes on a discretized 2D plane.…”

Section: Misbehavior Propertymentioning

confidence: 99%

“…The counting is usually closely related to the diversity property used. One can count the number of runs satisfying the conditions [53], [59], [60], [63], [68], [71] (denoted as All). Alternatively, one can base on subjective judgement [44], [57], [70], [72], [73], [74] (denoted as Subjective); distinct spatial and speed trajectory of the ego car [67] (denoted as Trajectory); distinct scenario parameters [45], [46], [49], [61] (denoted as Input); distinct scenario parameters regions [55] (denoted as Input Region); distinct scenario parameters with certain distance among them [50], [65] (denoted as Input Distance); or scenario parameters dispersion [66] (denoted as Input Dispersion).…”

Section: More Qualified Scenariosmentioning

confidence: 99%

“…No [47], [49], [63], [71], [72] Simulation Based Off-line Low [43], [44], [45], [46], [48] [50], [51], [52], [54], [55] [56], [57], [58], [60], [61] [62], [64], [65], [66], [67] [68], [70], [73], [74] Step Based Online High [53], [59], [69] TABLE 5: Comparison of different algorithm categories.…”

Section: Non-adaptivementioning

confidence: 99%

See 4 more Smart Citations

A Survey on Scenario-Based Testing for Automated Driving Systems in High-Fidelity Simulation

Zhong¹,

Tang²,

Zhou³

et al. 2021

Preprint

View full text Add to dashboard Cite

Automated Driving Systems (ADSS) have seen rapid progress in recent years. To ensure the safety and reliability of these systems, extensive testings are being conducted before their future mass deployment. Testing the system on the road is the closest to real-world and desirable approach, but it is incredibly costly. Also, it is infeasible to cover rare corner cases using such real-world testing. Thus, a popular alternative is to evaluate an ADS's performance in some well-designed challenging scenarios, a.k.a. scenario-based testing. High-fidelity simulators have been widely used in this setting to maximize flexibility and convenience in testing what-if scenarios. Although many works have been proposed offering diverse frameworks/methods for testing specific systems, the comparisons and connections among these works are still missing. To bridge this gap, in this work, we provide a generic formulation of scenario-based testing in high-fidelity simulation and conduct a literature review on the existing works. We further compare them and present the open challenges as well as potential future research directions.

show abstract

Section: High-fidelity Simulatormentioning

confidence: 99%