edx Flashcards
(70 cards)
1
Q
130 000 people are killed in traffic, world wide, each year.
A
False
2
Q
Cyclists, pedestrians and motorized 2-3 wheelers are typically considered vulnerable road users.
A
True
3
Q
The point-of-no-return is the point when only a safety system would be able to avoid the crash – not the human.
A
False
4
Q
High-risk driving is only about what the driver does.
A
False
5
Q
Which of the following are typical ways to study/understand what causes crashes in traffic?
a) Test-track experiments
b) Studying everyday driving
c) Investigating near-crash causation
d) Virtual testing
e) Analysis of real-crashes
f) Bench-testing
A
b, c, e
6
Q
Police-collected crash data is used to derive national statistics
A
True
7
Q
In-depth crash data typically includes video recording of crashes
A
False
8
Q
Limited data access can only be a problem for statistical data, but not for in-depth crash data
A
False
9
Q
In-depth crash data contains relatively few cases and a lot of information for each case
A
True
10
Q
Data harmonization is a major challenge when constructing multinational crash databases
A
True
11
Q
Intermediate level crash data contains information about driver age and gender
A
True
12
Q
Police-collected crash data contains deformation measurements for the vehicles involved in crashes
A
False
13
Q
Driving speed is typically present in in-depth data but not in police-reported data
A
True
14
Q
Posted speed limit is typically present in in-depth data but not in police-reported data
A
False
15
Q
Crash data can be used to identify road user groups that are most affected by road crashes
A
True
16
Q
Crash data can contribute to retrospective safety benefit evaluation
A
True
17
Q
Macroscopic crash data is typically collected by independent research teams
A
False
18
Q
Consider two different databases, A and B. Database A is based on crashes with personal injuries. Database B is based on tow-away crashes. If the crash data from the two databases is to be combined, which crash case filtering should be used?
a) Select all crashes from A.
b) Select all crashes from B.
c) Select only those crashes from A that involves tow-away
d) Select only those crashes from B that involves personal injury.
A
c, d
19
Q
Suppose a data set involving bicycle crashes is to be analyzed. The researchers suspect underreporting might be an issue when analyzing data. More specifically, what could those issues be?
a) Certain types of bicycle crashes are not reported.
b) Both single bicycle crashes and bicycle crashes involving pedestrians occur in the data.
c) Very few cases involve fatal injury.
d) Only a fraction of all crashes involving bicycles are reported.
A
a, d
20
Q
On-scene crash investigation is conducted closer in time to the crash than retrospective crash investigation.
A
True
21
Q
Crash reconstruction is mainly focused on understanding vehicle kinematics in the post-crash phase.
A
False
22
Q
Crash causation analysis with the DREAM method is aimed at identifying all factors that may have contributed to the crash.
A
True
23
Q
Which one of the following statements is true?
a) Drivers are responsible for more than a million crashes every year.
b) Driver impairments include physical disabilities.
c) If a system is able to recognize a safety critical situation, then it is an active safety system.
d) Active safety operates up to 1 s before the crash.
A
a
24
Q
Which of the following statements are true?
a) In active safety systems, threat assessment and decision making algorithms are used to recognize a critical situation and decide interventions.
b) Frontal collision warning is a lateral support system.
c) Today’s active safety systems may enable low-level automated driving.
A
a, c
25
Which one of the following descriptions matches the term data element?
a) a series of actions that you take in order to achieve a result
b) a piece of planned work or an activity that is finished over a period of time and intended to achieve a particular purpose
c) a device which is used to record that something is present or that there are changes in something
d) a piece of information, typically a fact or a number, collected to be examined and considered and used to help decision-making
e) a set of connected things or devices that operate together
d
26
Which one of the following descriptions matches the term sensor?
a) a series of actions that you take in order to achieve a result
b) a piece of planned work or an activity that is finished over a period of time and intended to achieve a particular purpose
c) a device which is used to record that something is present or that there are changes in something
d) a piece of information, typically a fact or a number, collected to be examined and considered and used to help decision-making
e) a set of connected things or devices that operate together
c
27
Which one of the following is an appropriate sensor specification for an accelerometer used for measuring driver brake responses and vehicle control?
a) Range of +- 40g
b) Resolution of 0.1g
c) Manage temperatures > 110 degrees
d) Range +-10m/s^2
c
28
Which of the following statement are true related to the use of intra-vehicle sensors?
a) Using different units across analysis, or within a company or project, for a particular sensor measure, can result in problems.
b) Under some circumstances constant rotation (e.g. turning), can produce erroneous estimates of linearly increasing lateral speed.
c) When analyzing brake pressure it is always obvious what is the source of the pressure data.
d) All sensors measuring rotation directly provide direction of rotation intrinsically.
e) Accuracy is the same as precision.
a, b
29
Suppose two accelerometer have the following specifications:
Accelerometer A:
- Full scale: ±100 m/s2
- Resolution: 10 bits
Accelerometer B:
- Full scale: ±40 m/s2
- Resolution: 16 bits
Calculate the resolution (m/s2) for accelerometer A.
0.195m/s2
The full scale is 2 · 100 = 200 m/s2
This full scale interval is matched to the number 210 = 1024
Hence, the resolution in the unit to be measured is 200 / 210 = 200 / 1024 = 0.195 m/s2.
30
Suppose two accelerometer have the following specifications:
Accelerometer A:
- Full scale: ±100 m/s2
- Resolution: 10 bits
Accelerometer B:
- Full scale: ±40 m/s2
- Resolution: 16 bits
Calculate the resolution (m/s2) for accelerometer B.
0.0012
The full scale is 2 · 40 = 80 m/s2
This full scale interval is matched to the number 216 = 65536
Hence, he resolution in the unit to be measured is 2·40 / 216 = 80 / 65536 = 0.0012 m/s2.
31
Which of the following statements are true related to differential GPS?
a) Differential GPS is using a support system of a moving reference station that has a known position.
b) Differential GPS cannot remove multipath errors but removes most of the ionospheric and tropospheric errors.
c) Network RTK improves the positioning by collecting information from several static reference GPS-receivers’ information into a common error correction for the specific position of the Network RTK GPS.
d) Onboard clock errors cannot be corrected by differential GPS.
b, c
32
Which one of the following statements about LIDAR is true?
a) The 3D (360 degree) LIDAR sensor can advantageously be placed in the front right bumper of a vehicle.
b) When receiving multiple reflections from a LIDAR using the first reflection is best.
c) It is not equally easy to identify a pedestrian in a 3D scanning LIDAR as with a 2D scanning LIDAR
c
33
If a 3D scanning LIDAR have only one pulsing LIDAR measurement unit, but scans 128 vertical lines with 0.1 degree resolution, and with 20 revolutions (full 360 degree) revolution per second, the scanner would have to have produce approximately which of the following individual measurements counts per second?
9 216 000 measurements/s
0.1 degree resolution and 20 revolutions per second means 360/0.1·20 measurements per revolution. This needs to be made 128 times a second, resulting in 360/0.1·20·128 = 9 216 000 measurements per second.
34
LIDAR: If the time from emission to reflection is 440 nanoseconds, what is the estimated distance (m) to an object? The speed of light is approximately 300 000 000 m/s.
66m
2 · D = c · t
D: Distance
c: Speed of light
t: time
D = c · t / 2 = 3·108 · 440·10^-9 / 2 = 66 m
35
Given that the reflectivity is good enough, how long time (micro-s) does it take from emission to reception of a LIDAR signal if an object is 300 m away?
2 microseconds
2 · D = c · t
D: Distance
c: Speed of light
t: time
t = 2 · D / c = 2 · 300 / 3·108 = 2microseconds
36
Disregarding limitations in number of bits available to represent it, what is the angular resolution if a 2D LIDAR sweeps a full revolution 12 times a second and LIDAR measurements occur 7200 times a second?
0.6 degrees
12 revolutions per second · 360 degrees per revolution / 7200 measurements per second = 0.6 degree resolution
37
Consider a car instrumented with a single vertical LIDAR beam plane having another car driving right in front of it at a distance of 150 m. Assume a completely straight road and that the lidar plane is completely horizontal (when not pitching). Calculate the difference in height of where the beam strikes if the instrumented car pitches with 0.5 degrees.
1.3m
h = L · tan(alpha) = 150 · tan(0.5) = 1.3 m
38
Which one of the following statements is true with respect to radar?
a) Relative speed is typically measured using the change of frequency of the radar pulse between when it is sent and received.
b) Modern radars typically use a fixed frequency.
c) The use of radar sensors in automotive applications has reached its peak.
d) The radar cross section is basically independent of frequency.
a
39
Which one of the following statements is true with respect to CMOS vs CCD sensors and vision sensor specifications?
a) Dynamic range describes how far away distances can be calculated with a vision sensor.
b) If the ego-vehicle just came to the top of a crest/hill (at the very top) and a mono camera was used to measure the distance to a lead vehicle, only using the flat-earth assumption method, the violation of the flat-earth assumption in range measurements would produce an overestimation of the range.
c) Interlace effects from interlace cameras are easily corrected in post-processing.
d) CMOS sensors pixels are intrinsically less uniform than CCD.
d
40
Which of the following statements are true regarding the near-infrared cameras and the difference between near-infrared and far-infrared cameras?
a) Near-infrared cameras typically require active illumination, while far-infrared cameras are passively recording “heat”.
b) Near-infrared cameras are more sensitive to blooming and glare from other vehicles than far-infrared are.
c) Near-infrared cameras cannot be used for stereo vision applications.
d) Near-infrared cameras typically provide a color image.
a, b
41
Which one of the following statements about the difference between time-of-flight cameras and FIR cameras is correct?
a) Time-of-flight cameras actively illuminate the scene while FIR cameras are passive sensors.
b) Time-of-flight cameras have worse performance at night than during daytime, while the opposite is true for FIR cameras.
c) FIR camera images look similar to NIR camera images.
d) Time-of-flight cameras are typically good for applications with objects at distances ranging from 15-50m.
a
42
Which are the correct statements about the difference between LIDARs, time-of-flight cameras and RADARs?
a) Both the time-of-flight camera and LIDAR transmit light, typically near-infrared.
b) LIDAR range is limited by the reflectivity of objects the transmitted signal encounters.
c) For both LIDARs and RADARS, speed can be calculated with the Doppler effect.
d) The angular resolution for most scanning LIDARs is better than most RADARs.
e) LIDARs have worse weather (rain and snow) performance than RADARs.
a, b, d, e
43
Which of the following statement are true?
a) Automated driving promises to make our roads safer by reducing human error.
b) Deciding the destination of a trip is not part of the dynamic driving task.
c) The dynamic driving task is part of object and event detection and recognition.
d) When an automated feature, such as ACC, exceeds its operational design domain, the driver taking over is the fallback.
a, b, c
44
Which of the following statement are true?
a) SAE levels are assigned based on DDT, OEDR, fallback, and ODD.
b) A feature providing L3 automation would be able to reach a minimal risk condition when exiting its ODD even if the driver would not take over.
c) For a single vehicle to reach L5, all vehicles in traffic must be automated.
d) Current active safety may provide L1 or L2 automation. An example is ACC providing L1 automation since many years back.
a, d
45
Which of the following statement are true?
a) A system can be cooperative even without wireless communication.
b) Cooperative safety systems may be active safety systems.
c) DSRC is a range of frequency reserved for communication among vehicles and between vehicles and the infrastructure.
d) When a truck exchange information wirelessly with a car, we call it V2I communication.
b, c
46
Which of the following statements are true?
a) Standards for V2V and V2I communication, such as DSRC, are necessary at all OSI layers to enable cooperative applications.
b) An LDM is a common database which each vehicle can assess wirelessly.
c) Data security is crucial for cooperative applications because hacking may make cooperative application unsafe.
d) Platooning requires that at least 80% of the cars in the platoon are equipped with wireless communication.
a, c
47
Please list eight of the current challenges for cooperative applications.
- critical mass
- wireless transmission
- positioning system
- standards for communication
- data management
- security
- privacy
- ownership
48
Which of the following statements are true?
a) Large naturalistic driving studies provide big data.
b) In a naturalistic dataset, data are seldom missing.
c) Naturalistic data can show the genuine driver behavior just before a crash occurs.
a, c
49
Which of the following statements are true?
a) According to the Heinrich triangle, severe accidents are more prevalent than less severe accidents.
b) The Heinrich triangle suggests that there may be a relation between crashes and near-crashes.
c) Near-crashes are often used as surrogates for crashes in analyses of naturalistic data.
b, c
50
Which of the following statements are true?
a) Naturalistic data may support the evaluation of automotive technologies such as active safety by enabling counterfactual analyses.
b) Naturalistic data is increasingly available.
c) Automated vehicles may become a new source of naturalistic data.
a, b, c
51
Which of the following behaviours represent an example of driver behavior?
a) A driver brakes and makes signs to a group of pedestrians to invite them to pass on the zebra crossing.
b) A driver opens the door of the vehicle when the engine of the vehicle is switched off.
c) A driver takes away one hand from the steering wheel to reach the phone and diverts the gaze from the road to the mobile phone to read a text message.
d) A driver drives in reverse gear and, during the backing maneuver, hits a motorcycle parked on the side of the road.
a, c, d
52
According to previous research, the human is a contributing factor in about...
90% of the crashes.
53
Mark is driving a vehicle equipped with Autopilot and the system is activated. He is approaching a signalized T intersection with a perpendicular leg on the right and he has right of way. Just before the intersection, there are some roadworks on the opposite lane. When Mark’s vehicle reaches the intersection, another vehicle stopped at the right perpendicular leg starts moving.
Which one of the following of Mark’s behaviours is appropriate during the driving with Autopilot when he reaches the intersection?
a) Keep his hands on the steering wheel and look at the roadworks on the opposite lane.
b) Keep his hands away from the steering wheel and text on his mobile phone.
c) Keep his hands away from the steering wheel and look at the road ahead.
d) Keep his hands on the steering wheel and check the movement of the other vehicle.
d
54
What are the main five factors influencing driver behaviour?
1. Individual characteristics
2. Driver motives
3. Socio-cultural values and norms
4. Technological factors
5. Driving context
55
Which of the following statements are correct?
a) Driving context can influence driver behaviour by limiting the driving speed.
b) The use of active safety systems (e.g. Adaptive Cruise Control) has been shown not to have an influence on driver’s behavior.
c) National differences has been shown not to have an influence on driver’s behavior.
d) Driver’s gender has been shown to have an influence on driver’s behavior.
a, d
56
Which of the following statements are correct?
a) Inhibitory motives are motives that encourage a person to achieve a specific goal.
b) The urge to reach a destination in time could be an example of excitatory motives.
c) The negative emotions experienced during a car crash could be an example of inhibitory motives.
d) Excitatory motives are motives that induce a person to avoid risks.
b, c
57
Which of the following statements about the Human Machine Interface (HMI) of the Adaptive Cruise Control (ACC) are correct?
a) The set speed – shown on the dashboard – is an example of information provided by the HMI.
b) The set time headway – shown on the dashboard – is an example of control provided by the HMI.
c) The button to set the time headway – located on the dashboard – is an example of control provided by the HMI.
d) The ACC status – shown on the dashboard – is an example of information provided by the HMI.
a, c, d
58
Which one of the following statements about the user-centered design process is correct?
a) The understanding of the needs of target users is not relevant for the user-centered design process.
b) The user-centered design process mainly aims to create stylish products.
c) Heuristic evaluations are not conducted during the user-centered design process.
d) The user-centered design process might require several iterations before completing the final product.
d
59
Which of the following ones could be the output of the driver model, indicated by the red question mark in the figure above?
a) Driver’s depression of the brake pedal (continuous variable: 0% - 100%)
b) Driver’s depression of the accelerator pedal (continuous variable: 0% - 100%)
c) Driver’s rotation of the steering wheel (continuous variable: 0 - 30 degrees)
d) Driver’s activation of Adaptive Cruise Control (binary variable: 0 - 1)
c
60
Which of the following ones are examples of driver models?
a) A model that describe the braking behaviour of drivers in rear-end emergency situations.
b) A model that describe the lateral distance maintained by drivers towards cyclists during an overtaking maneuver.
c) A model that describes the longitudinal behaviour (speed and time headway from the lead vehicle) of the Adaptive Cruise Control (ACC) system.
d) A model that describe drivers’ reaction times (time elapsed from the issue of a warning until driver depresses the brake pedal) after the warning issued by a Forward Collision Warning (FCW).
a, b, d
61
Angie is driving her car on a rural road at 60 km/h. Another car is travelling ahead at 40 km/h. The width of the lead vehicle is 1.8 meters. The distance between Angie’s car and the lead vehicle is 30 meters.
What is the approximate value of half of the optic field of view
Φ originated by the lead vehicle?
Φ=R/Z
R: Half the width of the vehicle
Z: Distance between eye and back of the vehicle
Φ: Half optic field of view
Φ = 0.9/30 = 0.03 radians
Φ = 0.03 * 180 /pi = 1.7 degrees
62
Please list the three main types of HMI modalities.
- auditory
- haptic
- visual
63
Which of the following statements are correct?
a) The visual HMI modality provides warnings to the drivers by emitting visual information and sound.
b) Visual HMI warnings can be shown in the instrument cluster or on the windscreen.
c) The vibration on the steering wheel is an example of haptic HMI warning.
d) Speech warning is an example of auditory HMI warning and it can be described as a pure tone or a combination of tones which has the objective of grabbing drivers’ attention.
b, c
64
Which of the following statements are correct?
a) Auditory HMI warnings can be used to provide continuous lower-priority information.
b) Visual HMI warnings can be used to grab drivers’ attention when drivers may be looking away.
c) Auditory HMI warnings can be used to indicate the onset of a system malfunction or limitation.
d) Haptic HMI warnings can be used to provide information if an auditory warning is unlikely to be effective.
c, d
65
Suppose that our goal is to analyze the effect of a variable in a crash database on the injury severity of the driver. Which of the following statements are correct in this context?
a. A confounder is a variable whose value affects both the value of the analyzed variable and the injury severity of the driver.
b. Only those variables can be confounders that are present in the analyzed database.
c. One way to address confounding is to divide the sample in groups in which confounding variables take similar values.
d. Age and gender of the driver may need to be investigated as potential confounders.
a, c, d
66
How does the sampling of fatal crashes only affect the estimation of fatality risk?
a. Fatality risk may be underestimated.
b. The estimate of fatality risk is not affected by the sampling criteria.
c. Fatality risk may be overestimated.
d. Under- or overestimation of fatality risk are equally likely.
c
As all sampled crashes include at least one fatally injured person, the relative frequency of fatalities in the sample may be way higher than in the general crash population, making it very likely that fatality risk is overestimated.
67
Impact speed is an outcome metric for both crashes and near-crashes.
False
68
Post-encroachment time (PET) is only a counterfactual (potential) severity metric.
False
69
Driver glance behaviors' impact on safety can’t be evaluated through counterfactual simulations.
False
70
Post-encroachment time (PET) is a relevant outcome metric that can be used in both crashes and near-crashes.
False
