Cognitive Systems Engineering Opens the Door to New Possibilties | Layla Akilan | TEDxDayton

TEDx Talks
28 Feb 202309:16

Summary

TLDRThis video explores the impact of bad design, using the Three Mile Island nuclear accident as a prime example. Poor system design, including confusing alarms and misleading data, contributed to the disaster. It highlights the need for user-centered design and cognitive systems engineering to improve human-machine interaction. By studying how humans think and work, we can create better systems that support decision-making, prevent errors, and enhance human performance. The video calls for a broader perspective in technology design, focusing on both users and the systems they interact with.

Takeaways

  • πŸ€” Bad design is a common issue in everyday technology, causing frustration and inefficiency.
  • ⚠️ The 1979 accident at Three Mile Island nuclear plant was a significant example of how poor design can lead to catastrophic consequences.
  • 🚫 A stuck open valve, along with confusing control room interfaces, led to operators misinterpreting the situation during the Three Mile Island incident.
  • πŸ’‘ The accident at Three Mile Island was not due to human error but rather a result of poor system design choices like misleading data and confusing alarm systems.
  • πŸ‘₯ User-centered design aims to overcome bad design by understanding users' mental models, addressing their pain points, and including them in the design process.
  • 🧠 As technology evolves, users are required to do less physical work but more cognitive tasks, like overseeing automated systems.
  • πŸ›  Cognitive Systems Engineering (CSE) focuses on improving the design of systems to better support users' cognitive demands in complex environments.
  • πŸ” CSE promotes studying users, their work, and the technology they interact with as part of a joint cognitive system to create more effective human-machine interactions.
  • πŸ”„ Effective system design considers the broader work domain and cognitive demands, as shown by how better control room interfaces could have mitigated the Three Mile Island incident.
  • πŸšͺ CSE encourages expanding perspectives to look beyond just users, emphasizing the interaction between people, technology, and work to improve overall human performance.

Q & A

  • What is the significance of bad design in the context of technology?

    -Bad design in technology, whether in everyday devices or complex systems, can lead to confusion, inefficiency, and even dangerous situations, as it creates obstacles for users by making systems harder to use and understand.

  • What was the major incident mentioned in the transcript related to nuclear power plant design?

    -The major incident was the Three Mile Island nuclear power plant accident on March 28, 1979, in Pennsylvania, where a series of design flaws and system failures led to the most significant nuclear power accident in U.S. history.

  • What triggered the initial failure in the Three Mile Island incident?

    -The failure began when a system malfunction prevented water from reaching the steam generators responsible for cooling the reactor core. This led to a temperature increase and the reactor shutting down automatically.

  • Why did the relief valve issue worsen the situation during the Three Mile Island accident?

    -Although the relief valve was designed to open to release pressure, it failed to close once pressure normalized, leading to cooling water pouring out. Control room operators mistakenly thought the valve was closed, worsening the situation.

  • How did poor design choices in the control room contribute to the Three Mile Island accident?

    -Poor design choices such as incorrect color coding, misleading data visualizations, and confusing alarm systems prevented operators from diagnosing the problem effectively, leading to misinformed decisions.

  • What is user-centered design, and how does it address the issues of bad design?

    -User-centered design focuses on empathizing with users, understanding their mental models, addressing their pain points, and involving them in the design process to create more intuitive and user-friendly systems.

  • How has the role of humans evolved with the advancement of technology, according to the transcript?

    -Initially, machines extended human physical abilities, but with modern interfaces and automation, humans have shifted from controlling machines to overseeing complex systems and processes, increasing cognitive work like problem-solving.

  • What is the focus of cognitive systems engineering as described in the transcript?

    -Cognitive systems engineering focuses on designing systems that better support human cognitive processes by studying how people think, interact with technology, and accomplish tasks within complex environments.

  • What is a 'joint cognitive system,' and why is it important in system design?

    -A joint cognitive system describes the interaction between people, technology, and work. Its essential properties come from how these elements interact, so studying users in isolation misses key insights into the system as a whole.

  • How could cognitive systems engineering have helped prevent the Three Mile Island accident?

    -Cognitive systems engineering could have helped by offering insights into designing a control room that properly supported operators, helping them better diagnose problems and make informed decisions during unexpected events.

Outlines

00:00

πŸ’‘ The Importance of User-Centered Design in Complex Systems

This paragraph discusses how examples of bad design, both simple and complex, can have severe consequences. It highlights the 1979 Three Mile Island accident as a case where poor design choices led to confusion among operators, resulting in a nuclear power plant failure. The text emphasizes that user-centered design emerged as a response to such issues, by focusing on empathy, understanding user mental models, and involving them in the design process. It stresses the importance of supporting human cognition as technology advances and highlights the role of cognitive systems engineering in creating effective human-machine interfaces.

05:02

πŸš— Joint Cognitive Systems: Understanding Human-Machine Interaction

This paragraph introduces the concept of a 'joint cognitive system' using a car as a relatable example. It explains that just as a car's essential properties are derived from the interaction between its mechanical parts, a joint cognitive system's properties come from the relationship between people, technology, and work. The text argues for the importance of studying users in the broader context of their interaction with technology, advocating for cognitive systems engineering as an approach to designing systems that enhance human performance by considering the entire work domain, user cognition, and technology in tandem.

Mindmap

Keywords

πŸ’‘Bad design

Bad design refers to systems or interfaces that are difficult for users to understand or operate, leading to frustration or errors. In the video, bad design is illustrated through everyday examples and more serious instances, such as the Three Mile Island nuclear accident, where poor design choices in the control room led to confusion during a crisis.

πŸ’‘Three Mile Island

The Three Mile Island accident in 1979 is a critical example used in the video to demonstrate the consequences of bad design at a large scale. Poor system design in the control room, such as incorrect color coding and misleading visual displays, contributed to operators’ inability to properly manage the failure, leading to a nuclear crisis.

πŸ’‘User-centered design

User-centered design is an approach that prioritizes understanding and addressing the needs, mental models, and pain points of users. The video emphasizes that this design philosophy emerged as a solution to the kind of bad design seen in systems like Three Mile Island, aiming to make interfaces more intuitive and effective.

πŸ’‘Cognitive systems engineering

Cognitive systems engineering (CSE) is a discipline focused on improving the interaction between humans and technology by designing systems that support human cognitive work, such as problem-solving and decision-making. The video highlights this approach as essential in addressing the complex demands placed on users by modern technology, especially in high-stakes environments like aviation and power plants.

πŸ’‘System failure

System failure refers to the breakdown or malfunction of a component within a larger system. The video uses the failure of the relief valve at Three Mile Island as an example, where the failure to close properly led to a loss of cooling water and an escalating nuclear crisis, exacerbated by poor interface design.

πŸ’‘Automation

Automation is the use of machines or software to perform tasks without human intervention. The video discusses how automation has shifted the role of humans from controlling machines to overseeing processes, such as pilots monitoring automated flight systems or drivers supervising autonomous vehicles. This change has increased cognitive demands on users, making effective system design even more critical.

πŸ’‘Joint cognitive system

A joint cognitive system refers to the interaction between humans, technology, and the work environment. The video explains that focusing solely on the user, without considering the broader system they are part of, limits understanding and effectiveness. Instead, cognitive systems engineering looks at how humans and technology work together as an integrated system.

πŸ’‘Work domain analysis

Work domain analysis is a method used in cognitive systems engineering to deeply understand the work, activities, and environment of users. The video suggests that such analysis could have helped in designing better control room interfaces at Three Mile Island, as it would have provided insights into the tasks operators perform and the challenges they face in complex scenarios.

πŸ’‘Human performance

Human performance refers to how effectively humans can carry out tasks or make decisions in a given system. The video stresses that the goal of cognitive systems engineering is to improve human performance by designing technology that supports users' cognitive abilities, preventing scenarios where poor design degrades their ability to perform, such as during the Three Mile Island incident.

πŸ’‘Mental models

Mental models are the internal representations that users form about how a system works based on their experiences and interactions. The video mentions how user-centered design seeks to align systems with these mental models to make them easier to use and understand, reducing confusion and errors in complex environments like control rooms.

Highlights

Simple examples of bad design are everywhere, and we all experience them through interactions with computers and devices.

On March 28, 1979, the most significant nuclear power plant accident occurred at Three Mile Island in Harrisburg, Pennsylvania.

A system failure at 4 AM prevented water from being sent to the steam generators, leading to an automatic reactor shutdown.

A stuck open valve caused cooling water to escape, exacerbating the situation, while control room operators were unaware due to misleading data.

The failure to diagnose and resolve the issue was not due to human error, but poor design choices in the control room such as misleading data visualizations and confusing alarms.

User-centered design emerged as a solution to poor design by focusing on user needs, mental models, and pain points.

Despite technological progress, systems remain complex and often hard to use, requiring better user support.

The evolution of the computer interface has shifted human roles from controlling machines to overseeing processes, increasing cognitive work.

Pilots and drivers now monitor automated systems, moving from direct control to overseeing processes.

Cognitive systems engineering is an evolving discipline that addresses the cognitive demands placed on users by modern technology.

According to systems thinking, a system's defining properties come from how its parts interact, not just from the individual components.

When designing for humans, it's important to consider the bigger system of interactions between people, technology, and work, known as a joint cognitive system.

Cognitive systems engineering requires a deep study of both the work domain and the cognitive work done by users.

This approach aims to design more effective human-machine teams by leveraging cutting-edge technologies like automation.

The ultimate goal of cognitive systems engineering is to improve human performance, rather than degrade it, as seen in the Three Mile Island incident.

Transcripts

play00:00

laughs

play00:03

[Music]

play00:08

[Applause]

play00:09

[Music]

play00:20

simple examples of bad design

play00:23

are everywhere

play00:25

we all experience them when we interact

play00:27

with our computers and our devices

play00:30

and they reveal how clumsy design

play00:33

can be really annoying

play00:36

more complicated examples of bad design

play00:39

reveal things that are harder to see

play00:41

like the consequences of poor design

play00:43

choices at larger scales

play00:46

on March 28 1979 in Harrisburg

play00:50

Pennsylvania the most significant

play00:52

accident to date in nuclear power plant

play00:55

history would occur

play00:56

in the unit 2 reactor at Three Mile

play00:59

Island

play01:01

it was 4 AM when the first system

play01:02

failure would arise

play01:04

and that failure would prevent water

play01:07

from being sent to the steam generators

play01:09

that were responsible for removing heat

play01:11

from the reactor core

play01:14

the resulting increase in temperature

play01:16

would cause the plant's reactor to

play01:17

automatically shut down

play01:20

in order to control the pressure that

play01:22

was now building in the primary system a

play01:25

relief valve was designed to open

play01:27

automatically

play01:28

and it did

play01:30

but when the pressure returned back to

play01:32

proper levels the valve should have

play01:34

closed and it didn't

play01:36

it became stuck open

play01:39

cooling water is now pouring out of the

play01:41

stuck open valve

play01:43

and to make matters worse

play01:45

control room operators thought that the

play01:48

valve was closed

play01:50

so alarms are going off and warning

play01:52

lights are flashing but no one can seem

play01:55

to make sense of what's Happening

play01:59

controllers who were responsible for

play02:01

monitoring the system were unable to

play02:04

diagnose the problem or make effective

play02:07

decisions about how to mitigate the

play02:09

situation

play02:11

and this wasn't because of human error

play02:14

this was the result

play02:16

of a series of poor design choices made

play02:19

in the control room like incorrect color

play02:22

codings misleading data visualizations

play02:25

incorrectly placed visual displays

play02:28

and confusing alarm systems

play02:32

user-centered design came along as a way

play02:34

to overcome bad design like this

play02:37

by empathizing with users uncovering

play02:40

their mental models addressing their

play02:42

pain points satisfying their needs

play02:45

and including them in the design process

play02:49

and while we have made progress

play02:52

we've learned a lot about users since

play02:54

then

play02:55

we all experience daily reminders that

play02:58

systems are still complex and hard to

play03:00

use

play03:03

we began using machines as a way to help

play03:05

us do physical work machines started as

play03:08

an extension of our physical abilities

play03:11

and then with early computers we started

play03:14

to use machines to do more advanced work

play03:16

computers helped us to perform complex

play03:18

calculations

play03:21

and now with the evolution of the

play03:23

computer interface

play03:24

it has once again shifted the role of

play03:26

human users from controlling machines

play03:30

to controlling processes

play03:32

resulting in less physical work

play03:34

but more cognitive work like problem

play03:37

solving and decision making

play03:39

Pilots they no longer just fly planes

play03:43

they oversee automated flight systems

play03:46

and a growing number of us no longer

play03:48

just drive cars

play03:50

we monitor our autonomous vehicles

play03:55

as a cognitive systems engineer I am

play03:57

fascinated in studying humans and how

play04:00

they interact with technology

play04:02

and ways in which we can design better

play04:05

system interfaces to support people in

play04:09

the way that we actually think

play04:13

the bottom line is technology is rapidly

play04:16

advancing and there's an increasing need

play04:19

to properly support users and the

play04:20

cognitive demands placed on them by new

play04:23

technology

play04:25

cognitive systems engineering is a

play04:28

discipline that arises from this need

play04:30

that's been evolving over the last 30

play04:32

years as an approach to Designing

play04:33

Technology

play04:35

so let's take a look at how this

play04:37

approach can not only help us better

play04:39

support users

play04:42

but extend the abilities of humans

play04:47

according to systems thinking Pioneer Dr

play04:50

Russell acoff a system has many parts

play04:55

that are all connected

play04:56

but what gives a system its defining

play04:59

properties is how the system comes

play05:01

together how the parts interact

play05:03

so let's use an example we can all

play05:05

relate to

play05:06

a car

play05:07

a car is a mechanical system has many

play05:10

parts

play05:12

but the car is the product of the

play05:15

interactions between the mechanical

play05:17

Parts the parts work together to

play05:21

transport you from one place to another

play05:23

and that is its essential property

play05:27

also reminds us that when you

play05:29

disassemble a system

play05:31

it loses its essential properties

play05:34

can no longer transport you

play05:37

when designing systems for humans

play05:40

focusing on the human is just one part

play05:44

of a bigger system

play05:47

and that system is called a joint

play05:49

cognitive system

play05:51

which describes the interaction between

play05:53

people technology and work

play05:57

and just like in our example of the car

play05:59

the essential properties of a joint

play06:01

cognitive system come from the

play06:03

relationships between the parts

play06:06

so we don't want to study users in a

play06:09

silo we want to consider users relative

play06:13

to this bigger picture because people

play06:17

are just one perspective into the

play06:19

problem

play06:20

and like any perspective

play06:23

can reveal some really important things

play06:25

while simultaneously hiding others

play06:29

expanding our perspectives on what

play06:32

matters by studying people in the way

play06:36

that they think

play06:37

in addition to the technology they

play06:40

interact with and the work that they

play06:42

need to accomplish

play06:44

allows us to design systems that support

play06:46

humans and their goals in meaningful

play06:48

ways

play06:50

a cognitive systems engineering approach

play06:52

requires breadth and depth

play06:55

so first we go abroad we study the work

play06:58

domain we gain a deep understanding of

play07:00

the work the people the activities the

play07:03

environment

play07:04

and then we go deep and we study the

play07:07

cognitive work we use cognitive analysis

play07:09

techniques to elicit Knowledge from

play07:11

users about how they think about their

play07:14

goals and how they respond in difficult

play07:16

scenarios how they adapt to surprise

play07:21

and finally

play07:23

we try to find ways to leverage

play07:25

cutting-edge technology like automation

play07:27

to re-envision the possibilities for how

play07:31

we can accomplish our work to create

play07:33

more effective human machine teams

play07:36

which can ultimately lead to humans

play07:38

being even more skillful in what they do

play07:42

because the goal

play07:44

is to improve human performance

play07:48

rather than creating systems that can

play07:52

degrade it like at Three Mile Island

play07:55

work domain analysis could have offered

play07:57

insights about how to design a control

play08:00

room to properly support operators

play08:03

and gaining a deeper understanding of

play08:05

the kinds of events that operators may

play08:07

experience and and how they might think

play08:09

through those

play08:11

could have lent itself to designing

play08:13

better system interfaces for diagnosing

play08:15

failures and providing decision support

play08:20

now is a turning point in technology

play08:24

a time for us to take a step back and

play08:26

reflect on our journey and our way

play08:29

forward

play08:31

if we fixate on users

play08:34

we will certainly see some things

play08:37

but we will miss others

play08:40

like staring through a keyhole

play08:43

but if we open the door with cognitive

play08:46

systems engineering

play08:48

it reveals A New Perspective one that

play08:51

suggests that

play08:53

people are just one piece of this puzzle

play09:02

[Music]

play09:05

foreign

play09:09

[Applause]

play09:10

[Music]

Rate This
β˜…
β˜…
β˜…
β˜…
β˜…

5.0 / 5 (0 votes)

Related Tags
Cognitive EngineeringHuman-TechnologyDesign FailuresNuclear PowerAutomationDecision MakingThree Mile IslandSystem DesignHuman PerformanceUser Experience