Exam 4

• 5 ironies of automation (1982):

  1. designers are human too! (mistakes will still be made)
  2. tasks that are not automated rely on humans, who are now less engaged and informed
  3. the human may need to take over if the system fails
  4. retrofitting automation can add complexity
  5. system skill shifts from one domain to another as automation
• Sheridan's First Level of Automation:

  1. no computer assistance; human must make all decisions and actions
• Sheridan's Levels of Automation ("decision aids"):

  2. computer offers a complete set of decision or action alternatives

  3. computer narrows the selection down to a few

  4. computer suggests one alternative
• Sheridan's Levels of Automation ("check systems"):

  5. computer executes that suggestion if the human approves

  6. computer allows the human a restricted time to veto before automatic execution
• Sheridan's Levels of Automation ("true automation"):

  7. computer executes automatically, then necessarily informs the human

  8. computer informs the human only if asked

  9. computer informs the human only if it decides to

  10. computer decides everything, and acts autonomously while ignoring the human
• Application of Automation (Parasuraman, Sheridan, and Wickens 2000):

  1. information acquisition (sensory processing)
  2. information analysis (working memory)
  3. decision selection (decision making)
  4. action implementation (response execution)
• Levels of automation (Society of Automotive Engineers):

  0. no automation - driver performs all tasks

  1. driver assistance - vehicle is controlled by the driver, but some driving assist features may be included in the vehicle design

  2. partial automation - vehicle has combined automated functions (ex. acceleration and steering), but the driver must remain engaged with the driving tasks and monitor the environment at all times
• Key Considerations in HF Automation Design:

  1. trust calibration - moderates overall interaction effectiveness; distrust (failure to trust automation when you should) vs. overtrust
  2. configuration errors - user programmed the automated system wrong
  3. workload - low workload phases are better; high workload is worse
  4. skills loss and training
  5. disrupted human-human interaction
• Simple tasks are easiest to automate, but maybe less helpful
• Changing levels can lead to mode awareness errors (ex. fighting autopilot errors)
• Levels of Automation Change Triggers:

  1. physiological (HR, GSR, EEG, eye data)
  2. critical events
  3. human performance (RT changes, exceeding established criteria and control input)
  4. models of task performance
• Adaptive Automation - system recommends or initiates changes in level of automation
• characteristics of adaptive automation:

  • change based on pre-established triggers
  • decreased acceptance of automation
  • decreased situation awareness
  • increase annoyance
  • decreased safety and performance
• adaptable automation - human operator requests or initiates change in level of automation
• characteristics of adaptable automation:

  • higher acceptance and trust
  • enhanced situation awareness
  • overloaded operators may not change level or may select an inappropriate level
• emerging concepts in automation:

  1. swarm automation - large numbers of simple semiautonomous devices perform tasks
  2. system-system interaction - how multiple automated systems interact
  3. transitionary period - ex. how a hybrid highway looks with both autonomous and non-autonomous vehicles operating together
  4. HF burden shifted from operation to design/development
• cognitive systems engineering:

  • addresses all aspects of a system (people, machines, policies, etc)
  • involves cognitive complexity (judgements, perceptions, reasoning, decision making, uncertainty)
  • requires team and cooperation/coordination
  • high stakes, time critical and dynamic environment
• three components of cognitive systems engineering:

  1. agents - people in the system, including the policy makers
  2. artifacts - displays, controls, signs, signals, communication
  3. world - environment, conditions, goal origins
• semiotics - study of signs and symbols, as well as their use or interpretation
• Rasmussen & CSE

  
• life cycle of CSE and tools:

  1. gather - observation, interview, survey, incident review
  2. analyze - HTA, CTA, requirement documentation, and user stories
  3. design - display/control design, prototyping
  4. evaluate - test systems or subsystems
• challenges in CSE:

  1. producing information that is actionable in the design process
  2. creation of artifacts that communicate needs/concerns between analysis and design
  3. assessing risk and managing tradeoffs
  4. traceability - trace design decisions back through the process with why questions
• 7 CSE design principles:

  1. support observability
  2. support predictability
  3. direct attention to critical info
  4. support exploration of solution space
  5. help users calibrate trust
  6. design to achieve common ground
  7. design to support goal attainment
• user experience and usability - general field of practice focusing on making consumer products more user friendly
• user experience in HF:

  • less emphasis on safety and performance
  • more emphasis on satisfaction and minimizing frustration
• usability - clear focus on learnability, efficiency, memorability, error prevention/recovery, and satisfaction
• skills needed for user experience career:

  1. research/analysis
  2. computer (prototyping/coding)
  3. design thinking
• design thinking (double diamond model):

  1. discover
  2. define
  3. develop
  4. deliver
• Discover (double diamond):

  • market research, user research, technology research, interviews, observation, empathy/journey mapping
• Define (double diamond):

  • analysis of gathered information
  • clustering, requirements gathering, project management plan
• develop (double diamond):

  • ideation, design charette/jam, design guidelines
  • sketch prototypes, formative testing
• deliver (double diamond):

  • high fidelity prototype
  • usability testing
  • revision planning
• 5 Stages of Design Thinking:

  1. empathize
  2. define
  3. ideate
  4. prototype
  5. test
• Aviation HF:

  • primary focus is safety
  • secondary is improved efficiency and effectiveness
• skills needed in aviation HF:

  1. research/analysis
  2. computer (prototyping/coding)
  3. design thinking
• concepts in aviation human factors:

  1. situation awareness and workload (ex. two pilots switching out on long flights)
  2. crew resource management (CRM) and team dynamics
  3. simulator research
  4. air traffic control
  5. flight environments (military, general, commercial, drones/unmanned aerial vehicles)
• aviation error prevention:

  1. "dirty dozen"
  2. originated in aviation maintenance
  3. emphasis is on promoting awareness and adherence to principles
• "dirty dozen" in aviation:

  1. lack of communication
  2. complacency
  3. lack of knowledge
  4. distraction
  5. lack of teamwork
  6. fatigue
  7. lack of resources
  8. pressure
  9. lack of assertiveness
  10. stress
  11. lack of awareness
  12. norms
• good problem statements:

  • clear and easy to understand
  • provides relevant context or background
  • is directive about goal but not the solution
  • broad enough for creative freedom but narrow enough to be manageable
  • human centered
• SCAMPER:

  • substitute
  • combine
  • adapt
  • modify
  • put to another use
  • eliminate
  • rearrange/reverse