Source

Woods, David, and Sidney Dekker. 2000. “Anticipating the Effects of Technological Change: A New Era of Dynamics for Human Factors.” Theoretical Issues in Ergonomics Science 1 (3): 272–82.

Article Source

TL;DR

Technological change impacts how people conceive their work, perform their duties and are evaluated against new expectations. Despite the acknowledged impact on “work as done”, Human Factors has historically been consulted late in the process, at the system evaluation phase. To be more effective, Human Factors needs to shift its focus to predicting how technological change will affect work so that it can better influence overall system design. Involving Human Factors in the design exploration phase is likely to produce better cooperative human-machine systems.

Introduction

Human Factors as a field has focused on describing how technology, people and organizations co-create work systems. One conclusion is that new technologies and capabilities don’t perfectly replace existing systems. Rather, they transform work and prompt new human adaptions. There is tempting oversimplification that new technology will perfectly replace existing human work without any other changes. This is called the substitution myth and while pervasive, has not proven to be accurate. (This is reminiscent of economics’ similar perfectly substitutable goods ).

The reality is that time-saving or capacity-expanding technologies often produce the opposite effects. From an analysis of Desert Storm:

Much of the equipment deployed…was designed to ease the burden on the operator, reduce fatigue, and simplify the tasks involved in operations. Instead, these advances were used to demand more from the operator.

Rather than a simple substitution, technological adoption passes through these stages:

• New roles are created
• There is a renegotiation of routine and exceptional
• What people consider “error” is accordingly shifted
• Different failure paths emerge in response to the changes

To maintain delivery and meet their individual goals, people innovate local adaptations to fill new gaps introduced by the new technologies. As production is expected to increase with the new capabilities, the elevated production pressures consume any newfound capacity. This acceleration is combined with new failure modes that emerge from the local adaptations to the new ways of working. This combination of factors helps define the Law of Stretched Systems. Systems are always pushed to their boundary of “safe” operation.

Adopting a black-box technology follows a predictable path:

1. New capabilities increase tempo and unintentional organizational coupling.
2. Unforeseen complexities emerge, induced by clumsy application of the new capabilities.
3. People doing the work make local adaptations to satisfy their goals.
4. Both the new complexities and local, unaccounted for adaptions are unintended side effects of the adoption.
5. Failures emerge because of changing conditions or incomplete adaptations.
6. Local adaptations are not disclosed to decision makers, who therefore attribute failures to human error.

In summary:

Under pressure from performance and efficiency demands, advances will be used to ask operational personnel to do more, do it faster or do it in more complex ways.

The conclusions is that capacity dividends are not applied to increasing safety margins or reducing tempo. They are applied to increase production levels (efficiency or throughput).

A new era of dynamics challenges traditional shortcuts

One problem for Human Factors is that at the time (2000), there was a tendency to produce reductive explanations that simplified analyses and could not keep pace with rapidly increasing technological change.

Woods and Dekker call for a new era in Human Factors, one focused on predicting the adaptations likely to emerge. This capability would enable Human Factors to engage earlier in the design process. Earlier engagement can help avoid the negative unintentional side effects of technological change.

They recommend Human Factors change their style of inquiry to maintain relevancy and increase impact. In the past, Human Factors experts were consulted only once an artifact or process was available. Providing feedback at only this stage is too late in the process. Human Factors value is not in evaluation, but in guiding exploration:

The problem in design today is not can it be built, but rather what would be useful to build given the wide array of possibilities new technology provides.

When Human Factors is consulted late in the design process, negative feedback is too easily dismissed because of the sunk cost of the existing design. Rapid prototyping is a perceived mitigation, but that only shortens the feedback cycle rather than changing the dynamic. What’s needed is to include Human Factors in the ideation stage.

Which is a problem, since predicting practitioner local adaptations creates an envisioned world problem. New technologies transform work, but those transformations are only speculative during the ideation phase. We can only guess about local adaptive behaviors due to imagined changes to working conditions.

A way out of this counterfactual is to leverage an iterative, experimental and empirical methodology. Human Factors should acknowledge that design concepts are hypotheses, test them, then evolve them. This type of design is not commonplace.

Predicting post-conditions of technology change

Assuming Human Factors changes to be more experimental and empirical, success will be measured by it’s ability to anticipate unintended effects and create new design options.

To produce these outcomes, Human Factors needs new tools:

• data
• models
• predictions
• design

Data can be produced by more case studies of organizational change that feed into model definition. Models are generalizations of the dynamics of change and adaptation in response to technological introduction. With models, Human Factors can be used to predict how technology changes “will transform roles, demands and activities, and where it will insert new vulnerabilities into the operational world.”

The ultimate goal is to use this feedback loop to prompt new design alternatives that increase human capacity, identify leverage points, and minimize undesirable side effects.

Achieving this ideal state is challenging because of the Envisioned World Problem. Basically, we don’t know what’s going to happen across at least two dimensions:

• plurality : future impacts are multi-valued
• underspecification: even with a single option, the effects are vague

Is it even possible to predict the full scope of future changes? Maybe, but those predictions suffer from two other limitations:

• ungrounded: envisioned concepts are not and cannot be backed by empirical results.
• overconfident: proponents can mistakenly assert that only the known predicted impacts will be realized.

There are different methods of investigations to compensate for these limitations (ethnological, functional, artifact, etc.). A powerful technique is the future incident technique, where a scenario is constructed from domain invariants (eg: maintaining flight path independence) and classic design failures (eg: clumsy automation or coordination surprises). Scenario designers game-play where given the envisioned world, an expected failure mode is injected and participants share their perspective.

[T]he use of concrete scenarios to anchor participants in the details of coordination communication, decision making and knowledge exchange necessary to handle the situation successfully is critical to this method. The incidents are illustrations of where a future architecture may be vulnerable or how it may break down, thus inviting practitioners and developers alike to think critically about the requirements for effective problem-solving in the envisioned world.

This approach moves Human Factors out of the verification phase and into the exploration and design phase. It’s most effective when (1) the problems to be solved in the test are representative of the actual world and (2) practitioner expertise is expressed in the discussion.

Conclusion

This paper is a call for a new type of Human Factors, one tailored for the new era’s rate of technological and organizational change. This new Human Factors “foundation for the future is observations of how technological and organizational changes transform cognitive and collaborative activities and demands, and how people in turn adapt to those changes.”

Attributions

Photo by Klim Musalimov on Unsplash