Investor Event Deck - RSS Jack Weast Sr. Principal Engineer, Intel - - PowerPoint PPT Presentation

investor event deck rss
SMART_READER_LITE
LIVE PREVIEW

Investor Event Deck - RSS Jack Weast Sr. Principal Engineer, Intel - - PowerPoint PPT Presentation

Investor Event Deck - RSS Jack Weast Sr. Principal Engineer, Intel VP, Automated Vehicle Standards, Mobileye How would you define driving safely for an AV? A statistical argument Self-driving cars should be statistically better than a


slide-1
SLIDE 1

Investor Event Deck - RSS

Jack Weast

  • Sr. Principal Engineer, Intel

VP, Automated Vehicle Standards, Mobileye

slide-2
SLIDE 2
slide-3
SLIDE 3

How would you define “driving safely” for an AV?

Self-driving cars should be statistically better than a human driver

A statistical argument

slide-4
SLIDE 4

Miles Driven

The more miles I drive without a crash, the safer I am Miles driven here Not the same as here

slide-5
SLIDE 5

Disengagements

Minimize the number of times the ADS fails and requires a takeover Why it’s insufficient Similar to miles driven, depends

  • n where & when

Incentive to avoid the tough environments likely to trigger disengagements

slide-6
SLIDE 6

How would you define “driving safely” for an AV?

The AV only needs to strictly obey the rules of the road

A rules-based argument

slide-7
SLIDE 7

Even if we strictly follow the rules, others around us may not do the same

Follow the Rules of the Road

STOP

slide-8
SLIDE 8

Even if we strictly follow the rules, others around us may not do the same

Follow the Rules of the Road

STOP

slide-9
SLIDE 9

How would you define “driving safely” for an AV?

Avoid collisions at all costs

A catch-all

slide-10
SLIDE 10

The AV Must Avoid Collisions at All Costs

slide-11
SLIDE 11

The AV Must Avoid Collisions at All Costs

slide-12
SLIDE 12

So what do we do?

slide-13
SLIDE 13

So what do we do?

What do humans do?

slide-14
SLIDE 14

Establish priority of road agent interests to avoid collisions

  • Come to complete stop

at red lights

  • Don’t cross a double-yellow line
  • Obey posted speed limits
  • Yield to other road users

when posted Set limits on vehicle operation

Explicit Traffic Rules

slide-15
SLIDE 15

A general set of principles applied by the driver

  • Keep a safe distance from the car

in front of you

  • Drive cautiously under limited

visibility

  • Don’t drive slow in the fast lane
  • Don’t cut off other drivers

Flexible, culturally dependent

Implicit Traffic Rules

slide-16
SLIDE 16

An open, transparent, technology neutral safety model for autonomous driving

Responsibility Sensitive Safety

RSS digitizes the implicit rules of the road, providing a check on AV decision-making, and a technology-neutral performance benchmark for regulators

slide-17
SLIDE 17

Implicit Rules of the Road

Essential for Navigating Complex Scenarios

slide-18
SLIDE 18

Rules of RSS

Rules to verify AV safety & performance

Do not hit someone from behind

1

Do not cut-in recklessly

2

Right-of-Way is given, not taken

3

Be careful in areas with limited visibility

4

If you can avoid a crash without causing another, you must

5

slide-19
SLIDE 19
slide-20
SLIDE 20
slide-21
SLIDE 21
slide-22
SLIDE 22
slide-23
SLIDE 23
slide-24
SLIDE 24
slide-25
SLIDE 25
slide-26
SLIDE 26
slide-27
SLIDE 27
slide-28
SLIDE 28
slide-29
SLIDE 29
slide-30
SLIDE 30
slide-31
SLIDE 31
slide-32
SLIDE 32
slide-33
SLIDE 33
slide-34
SLIDE 34

Does it work?

slide-35
SLIDE 35
slide-36
SLIDE 36
slide-37
SLIDE 37

Proactive Regulation of AV

An opportunity to get ahead of the curve

For the first time, we have the chance to define in advance and not after the fact, the desired balance of safety, utility, and efficiency

  • f AV’s on the road
slide-38
SLIDE 38

But there’s more to safety than math

slide-39
SLIDE 39
slide-40
SLIDE 40

Let’s Talk About “Safety”

Life & Death Safety

RSS

Psychological Safety

Trust

A Psychological sense of safety is based on human trust

slide-41
SLIDE 41

Why Trust Matters

“The majority of U.S. drivers seek autonomous technologies in their next vehicle, but continue to fear the fully self-driving car.” 75% “of U.S. drivers report feeling afraid to ride in a self-driving car.” “Psychology research shows people mistrust those who make moral decisions by calculating costs and benefits – like computers do”

April 24, 2017 “Why are we reluctant to trust robots?” March 7, 2017 “Americans Feel Unsafe Sharing the Road with Fully Self-Driving Cars”

slide-42
SLIDE 42

What Do We Mean by “Trust”?

“Firm belief in the character, strength,

  • r truth of someone or something.”

‒ Source: Merriam-Webster Dictionary Online

slide-43
SLIDE 43

The Study

slide-44
SLIDE 44

The Journey

  • REQUESTING A VEHICLE
  • STARTING A TRIP
  • MAKING CHANGES TO A TRIP
  • HANDLING ERRORS AND EMERGENCIES
  • PULLING OVER AND EXITING
slide-45
SLIDE 45

Contradictions = Tension

Participants expressed contradictory viewpoints in 7 key areas:

  • 1. Human judgment vs.

Machine judgment

  • 2. Personalized space vs.

Lack of assistance

  • 3. Make me aware vs.

Unburden me from being aware

  • 4. Giving up control of the vehicle vs.

Gaining new control of the vehicle

  • 5. How it works vs.

Proof it works

  • 6. Tell me vs.

Listen to me

  • 7. Following the rules vs.

Human interpretation of the rules

What Trust Should Look Like

slide-46
SLIDE 46

Introducing the RSS Concept Experience

Bridging the Gap Between Technical and Psychological Safety

slide-47
SLIDE 47

RSS HMI – An Embodiment of Human Trust in an In-Vehicle Experience

how it works VS. Prove it works

slide-48
SLIDE 48

RSS HMI – An Embodiment of Human Trust in an In-Vehicle Experience

slide-49
SLIDE 49

RSS HMI – An Embodiment of Human Trust in an In-Vehicle Experience

slide-50
SLIDE 50
slide-51
SLIDE 51