Are we taking the Human Factors Researcher out of the Loop? Sanna - - PowerPoint PPT Presentation

▶
are we taking the human factors
SMART_READER_LITE
LIVE PREVIEW

Are we taking the Human Factors Researcher out of the Loop? Sanna - - PowerPoint PPT Presentation

Feb 15, 2023 243 likes •669 views

Automated Driving Are we taking the Human Factors Researcher out of the Loop? Sanna Pampel GOTO Conference Copenhagen October 1 2016 GOTO Conference Copenhagen October 2 2016 Sanna Pampel Background in information systems and

slide-1
SLIDE 1

Automated Driving Are we taking the Human Factors Researcher out of the Loop?

Sanna Pampel

GOTO Conference Copenhagen October 2016 1

slide-2
SLIDE 2

GOTO Conference Copenhagen October 2016 2

slide-3
SLIDE 3

GOTO Conference Copenhagen October 2016 3

  • Background in information systems

and software development

  • PhD about drivers’ mental models in

Leeds

  • Today research Fellow for

Automotive Human Factors

Sanna Pampel

  • Automotive research in Nottingham

– Mirrorless cars – Emotions and experiences – HMI in partially automated driving

GOTO Conference Copenhagen October 2016

slide-4
SLIDE 4

GOTO Conference Copenhagen October 2016 4

  • ~25 staff, ~30 PhD students
  • Researching human beings and:

– Novel technologies – Complex systems – Safety – Implications on trust and user behaviour

Human Factors Research Group

  • Automotive HMI work – 5 staff, 8 PhD students

– Funded by vehicle manufacturers, EU, UK government

GOTO Conference Copenhagen October 2016

slide-5
SLIDE 5

Core Human Factors issues

Behavioural adaptation Attention Trust Workload and stress Situation awareness Human error Risk Anthropometry/ Biomechanics Mental models

GOTO Conference Copenhagen October 2016 5

slide-6
SLIDE 6

6

Automated Driving

Are we taking the Human Factors Researcher out of the Loop?

slide-7
SLIDE 7

British drivers spend 124 hours in traffic jams each year1

GOTO Conference Copenhagen October 2016 7

1 http://www.inrix.com

slide-8
SLIDE 8

Definitions

8

Automated

  • “automatically controlled operation of an apparatus, process or

system by mechanical or electronic devices that take the place of human labour” 1 Takeover and handover Takeover is initiated by driver and handover by the vehicle Autonomous

  • “not subject to control from outside; independent” 1
  • driverless

1 http://dictionary.reference.com

GOTO Conference Copenhagen October 2016

slide-9
SLIDE 9

Vehicle automation levels

9

Level 5 Full automation Level 4 High automation Level 3 Conditional automation Level 2 Partial automation Level 1 Assisted Level 0 No automation

Vehicle control Operator control

SAE document J3016, see http://standards.sae.org/j3016_201401 GOTO Conference Copenhagen October 2016

slide-10
SLIDE 10

Vehicle automation levels

10

Level 5 Full automation Level 4 High automation Level 3 Conditional automation Level 2 Partial automation Level 1 Assisted Level 0 No automation

Vehicle control Operator control

SAE document J3016, see http://standards.sae.org/j3016_201401 GOTO Conference Copenhagen October 2016

slide-11
SLIDE 11

Level 0 No automation

By Rmhermen - Transferred from en.wikipedia

slide-12
SLIDE 12

Vehicle automation levels

12

Level 5 Full automation Level 4 High automation Level 3 Conditional automation Level 2 Partial automation Level 1 Assisted Level 0 No automation

Vehicle control Operator control

SAE document J3016, see http://standards.sae.org/j3016_201401 GOTO Conference Copenhagen October 2016

slide-13
SLIDE 13

Level 1 Assisted

By Nozilla - Transferred from en.wikipedia

slide-14
SLIDE 14

Vehicle automation levels

14

Level 5 Full automation Level 4 High automation Level 3 Conditional automation Level 2 Partial automation Level 1 Assisted Level 0 No automation

Vehicle control Operator control

SAE document J3016, see http://standards.sae.org/j3016_201401 GOTO Conference Copenhagen October 2016

slide-15
SLIDE 15

Level 2 Partial automation

By OSX - Transferred from en.wikipedia

slide-16
SLIDE 16

Vehicle automation levels

16

Level 5 Full automation Level 4 High automation Level 3 Conditional automation Level 2 Partial automation Level 1 Assisted Level 0 No automation

Vehicle control Operator control

SAE document J3016, see http://standards.sae.org/j3016_201401 GOTO Conference Copenhagen October 2016

slide-17
SLIDE 17

Level 3 Conditional automation

By Tesla Motors Inc. - Tesla Motors Inc.

slide-18
SLIDE 18

Vehicle automation levels

18

Level 5 Full automation Level 4 High automation Level 3 Conditional automation Level 2 Partial automation Level 1 Assisted Level 0 No automation

Vehicle control Operator control

SAE document J3016, see http://standards.sae.org/j3016_201401 GOTO Conference Copenhagen October 2016

slide-19
SLIDE 19

Level 4 High automation

By Grendelkhan - Transferred from en.wikipedia

slide-20
SLIDE 20

Vehicle automation levels

20

Level 5 Full automation Level 4 High automation Level 3 Conditional automation Level 2 Partial automation Level 1 Assisted Level 0 No automation

Vehicle control Operator control

SAE document J3016, see http://standards.sae.org/j3016_201401 GOTO Conference Copenhagen October 2016

slide-21
SLIDE 21

Level 5 Full automation

Door Maurits Vink - Transferred from en.wikipedia

slide-22
SLIDE 22

Vehicle automation state-of-the-art

  • Audi, Mercedes, BMW level 3
  • Ford ‘affordable’ Level 3
  • Tesla level 3

22 GOTO Conference Copenhagen October 2016

  • Google car level 3/4 with hundreds of

handovers recorded

  • Level 5 not for another 10 years1

Source: Tesla Motors Inc. Department for Transport: 'The pathway to driverless cars: summary report and action plan' (DfT, 2015)

slide-23
SLIDE 23

Level 3 requires a responsible ‘driver’

  • Driver engagement

– Fatigue and mindwandering – Long takeover times – Degradation of skills – Trust and complacency – Misuse

23

First fatalities in level 3 automation

GOTO Conference Copenhagen October 2016

slide-24
SLIDE 24
  • Re-engaging the driver

– Keeping the driver in the loop

24 GOTO Conference Copenhagen October 2016

– Encouraging manual operations – Measuring driver state

Level 3 requires a responsible ‘driver’

slide-25
SLIDE 25

Jump to full automation

25

Level 5 Full automation Level 4 High automation Level 3 Conditional automation Level 2 Partial automation Level 1 Assisted Level 0 No automation

GOTO Conference Copenhagen October 2016

slide-26
SLIDE 26

Jump to full automation

26

Level 5 Full automation Level 4 High automation Level 3 Conditional automation Level 2 Partial automation Level 1 Assisted Level 0 No automation

GOTO Conference Copenhagen October 2016

slide-27
SLIDE 27
  • Overcoming the intermediate levels
  • Tesla

– Fully autonomous by 2018?

  • Ford

– Fully autonomous by 2021?

27

Does this mean we do not need to take the human into account?

GOTO Conference Copenhagen October 2016

Jump to full automation

slide-28
SLIDE 28

Fully autonomous vehicles – no humans?

  • Human in vehicles
  • Humans outside of vehicles
  • Traffic environment
  • Society

GOTO Conference Copenhagen October 2016 28

slide-29
SLIDE 29

Humans in vehicles

  • Trust and adoption
  • Emotions, comfort and experiences
  • Communication with car
  • Non-driving activities
  • Driving pleasure

– Will people not be allowed to actively drive?

29 GOTO Conference Copenhagen October 2016

slide-30
SLIDE 30

Humans outside of vehicles

  • Safety
  • Trust and acceptance
  • Predictability and communication with ‘driver’

30 GOTO Conference Copenhagen October 2016

  • Transition and mixed

traffic

– Regional and cultural differences – Feedback into R&D

slide-31
SLIDE 31

Traffic environment

  • Ethical decisions
  • Impact on traffic safety,

travel quality, travel time, lifestyles

31

– Mobility improvement and inclusivity

  • Connected vehicles
  • Misuse of ‘safe’ cars in traffic

GOTO Conference Copenhagen October 2016

slide-32
SLIDE 32

Society

  • Laws, regulations
  • Unemployed taxi/bus/lorry drivers
  • Ownership and business models

– Mobility as a service

  • Security and insurance

32 GOTO Conference Copenhagen October 2016

slide-33
SLIDE 33
  • Maintenance with less hassle
  • Knowing driver state to enhance experience

33 GOTO Conference Copenhagen October 2016

Artificial Intelligence

  • Adapting to driver’s

goals

– Personalisation

slide-34
SLIDE 34

34

  • Learn human bad habits

– Google car accident

  • Learning mechanisms can

adapt over time

– Rules need to be put in place – Will humans understand the car’s behaviour changes?

GOTO Conference Copenhagen October 2016

Artificial Intelligence

Photograph taken by Polimerek in MIT Mus.

slide-35
SLIDE 35

Artificial Intelligence

  • Shall autonomous cars drive like humans or

machines?

35

– A better version of ourselves? – Disappointed inevitable

GOTO Conference Copenhagen October 2016

slide-36
SLIDE 36

Can we address the challenges?

  • Fully autonomous cars not existing yet

– Only in limited areas

  • Many questions

36

How can we find answers now?

GOTO Conference Copenhagen October 2016

slide-37
SLIDE 37
  • Explore problem
  • Find research question

37 GOTO Conference Copenhagen October 2016

  • Answer might not involve full scenario
  • More opportunities as cars become

increasingly automated

Human factors exploring solutions

slide-38
SLIDE 38
  • Automation promises safety and increased

road capacity

  • Partially automated vehicles are an exciting

human factors research field

  • However, engaging/re-engaging the driver is

difficult, and potentially dangerous

  • Hence, announcements of fully autonomous

vehicles

38 GOTO Conference Copenhagen October 2016

Take aways

slide-39
SLIDE 39
  • Human factors challenges remain

– Human in vehicles – Humans outside of vehicles – Traffic environment – Society

  • Interesting challenges posed by AI
  • Research efforts to tackle questions before

autonomous vehicles are widely in use

39 GOTO Conference Copenhagen October 2016

Take aways

slide-40
SLIDE 40

40

So,

are we taking the Human Factors Researcher out of the Loop?

No.

Humans will still need to be considered.

slide-41
SLIDE 41

GOTO Conference Copenhagen October 2016 41