the future role of the human meetings seminars and... · 2019-01-29 · 1 the future role of the...
TRANSCRIPT
1
The future Role of the Human Towards a human centred approach to
implementation of technology
ICAO EUR/NAT Vision workshop 29-30.1.2019, Paris
NAT2030 PR09
The future: ’We need to change our approach’
2
• More technology and connectivity • Trajectory based operations, Drones, Remote operations, virtual ANS, etc
Erik Strasser, 2018
Program
• Why technology? • Technology in ATM • Why doesn’t it always work? • What can we do?
• The Joint Human Machine System
• Values • Conclusion
3
4
Why Technology?
I, II, II+,…
Why Support this task with technology?
5
Often the desire to, is driven by: ”We can, therefore we will”
Technology in ATM
7
• Of course, technologies can be of enormous help.
• But we must remember that they often disappoint, especially their users and managers.
• We would feel more guilty of making such claims if the track record of technology was better.
• In practice, we recall that most investments in IT, for example, do not meet their objectives.
(Landauer, 1995; Clegg et al., 1997; Nickerson & Landauer, 1997.
Why Technology?
Learning from history?
8
ATM 2000
Flexible Use of Airspace 1996
1990
SES 2004 SES II 2009 SES II+ 2013
9
“Anyone surveying the air traffic control systems of today can observe that even the technology which has been available for many years is still far from being uniformly applied. Technically we know, for example, how to engineer automatic data transfer between air traffic control centres; such facilities are not on the “leading edge’” of advanced technology. Yet controllers in much of Europe are still making telephone calls to pass boundary estimates on individual flights. Radar was in use for the Second World War: yet European air traffic control is still, fundamentally based on procedural concepts: in some areas. radar cover does not even exist.”
1992 DG Eurocontrol Keith Mack:
Why doesn’t it always work? (as expected or promised?)
Boeing this week issued a global bulletin advising pilots to follow its operations manual in such cases. But to do so, experts said, would have required Flight 610’s captain, Bhavye Suneja, a 31-year-old Indian citizen, and his co-pilot, Harvino, a 41-year-old Indonesian, to have made decisions in seconds at a moment of near-certain panic.They would have had to recognize that a problem with the readings on the cockpit display was causing the sudden descent. Then, according to the F.A.A., they would have had to grab physical control of the plane. That would not have been a simple matter of pushing a button. Instead, pilots said, Captain Suneja could have braced his feet on the dashboard and yanked the yoke, or control wheel, back with all his strength. Or he could have undertaken a four-step process to shut off power to electric motors in the aircraft’s tail that were wrongly causing the plane’s nose to pitch downward.
New York Times, 9.11.2018
Coping with complexity
13
Examples of everyday adaptions
Why doesn’t it always work? (as expected or promised?)
CPDLC + FRA
Work as Imagined Work as done
Why do socio-technical
systems work?
• Highly automated systems (in part developed to overcome the weaknesses of humans) in practice may only work because the human agents are so flexible and such good copers that they can make anything work
• Humans are creative and resourceful (Norman, 1998)
• Humans are quick learners, take responsibility, can react on weak signals, can improvise, dedication and emotions
14
What can we do?
15
Two different mindsets:
”we, the designers of this new system, need to find ways of getting you, the end-users, to participate in its design''.
”we, the end-users of the system, need to find ways of getting you, the experts in technology design, job design, business processes (etc.), to participate with us in its design''.
CLEGG, M. GRAY AND P. E. WATERSON, 2000
Design is about ownership.
The Joint Human Machine System
16
Joint: Human and Machine
17
Human vs. Machine Human manages Machine
Or vice versa
Human and Machine Complementarity of Human and Machine
(Wäfler et al., 2003)
Function Allocation
18
Human vs. Machine Human manages Machine
Or vice versa
Human and Machine Complementarity of Human and Machine
Function
Competition
Function
✕
Cooperation Joint ✓
(Wäfler et al., 2003)
From MABA-MABA to Joint Teaming
19
Joint
OECD Everyday math
11% of adults better than AI
Japan: 37%
Sweden: 36%
(Elliott, 2017)
• Machines show amazing capability
• Still humans and machines are complementary
• Machine • Has deep knowledge • Calculates: May give good answers
• Human • Has cross-linked knowledge • Thinks: May ask good questions
(Brynjolfsson & McAfee, 2017; Floridi, 2015)
The Human Contribution
20
Joint
• Verification of technical decisions
• Improving technical system • Learning from technical system
(Brynjolfsson & McAfee, 2017; Faust, 2007; Floridi, 2015, Manzey, 2012; Samek et al., 2017; Shively et al., 2017, Wäfler at al., 2003)
• Manage uncertainty
• Take responsibility • Show commitment and
dedication
• … • Empathy, creativity,
improvisation
Preconditions
Clear human role in the system
Respective system design • Avoiding deskilling • Allowing continuous development
of expertise (80% tacit)
• Job design required • Ergonomics, UI/UX not enough
• Participatory approach a must
The Joint Human Machine System goal
21
Closer collaboration between operators, system designers, system decision makers and the public Human and machine compliment each other to achieve system goals A humanistic design that allows humans to recover from the rare high-risk scenarios
Automation should not be Human versus machine, automation should be seen as human-machine coordination as a team.
HALA, 2010
What is needed?
• Planning for the future - we have to prepare the current system with the persons in the system to the upcoming changes.
• Future technical improvements needs to focus on systems rather than components and on functions rather than structures to be able to cope with complexity
22