IEEE Technology and Society Magazine - December 2019 - 65

arena. Understanding the inflow/outflow is very important, as excessive queueing is not just an inconvenience, but potentially dangerous. The recent "We Are
FSTVL" [11] event saw attendees collapse as they
queued for hours in the sun, without water or shade.
However, while these models provide us with an
exceptionally powerful tool to investigate the design of
infrastructure, they suffer from several drawbacks. The
major drawback is a lack of uniform verification and validation methods that confirm whether a particular
model is sufficiently accurate in simulating a scenario.
This verification and validation has been performed for
fire evacuation [12], and efforts are ongoing to introduce
these standards for other use cases.
A further drawback of these models is the inherent
ethical concerns that should be recognized and built
into the design process.

Ethical Concerns
Each specific use case raises questions regarding the
intersection of user equity and design approaches.
When developing a model the designer needs to ask a
few questions about the users:
Who am I primarily designing for?
Who else should I reasonably be designing for?
■ Will this change in the future?
■
■

The designer will then begin to make a model
depending on the answers to these questions. The models themselves are only as good as the data used to calibrate them, and the complexity of the model
determines the different behaviors it is capable of predicting. To expand on this point, it is worth understanding that groups move differently from individuals.
Familial groups may move differently than social
groups, whereas typical two-person and three-person
groups move significantly differently than a group of
children on a school trip. These differences might be
expressed in movement speed, speed-density relationships, wayfinding approaches, or interpersonal distances. All of our understanding of how these groups move
is based on experimental and real-world data (for an
example dataset, see [13]).
These differences will significantly affect the design of
infrastructure projects. It can be seen quite clearly that
the occupants of a typical gym differ quite considerably
from that of a care home, and with those differences
come significant changes to the standard building
design. But there might also be a larger variance in occupant demographics, which presents issues to an engineer
attempting to specialize their designs.
So why consider a specific range of user profiles when
designing a new project, rather than taking an approach
DECEMBER 2019

∕

The pedestrian dynamics approach
can have huge impacts in the
day-to-day life of pedestrians,
as well as in emergency situations
as human behavior becomes
more unpredictable.

that defines an "average" user? Using a first-principles
approach, it is far better to understand how actual users
will behave, rather than making assumptions that give
rise to a "rules-of-thumb" approach. These assumptions
must be subsequently justified, and result in a less flexible design. By putting the onus on the engineers to
design buildings for specified demographics, the movement of people will be much better understood, and the
resulting building design will far more efficient. This
informed, nuanced analysis will result in a user experience that is better, safer, and more tailored.
However, it is important to note that when a designer
specifically focuses on one group of people in their
model, they explicitly omit other groups. These other
groups may have a significant effect on the design produced, and while it is possible to design for multiple
groups, including for projected future demographic
changes, this quickly becomes an optimization problem
with a poorly defined cost-function. Unable to fully
accommodate specific groups of people, an engineer
might rely on proscriptive codes, and the design process therefore becomes at least partially determined by
experience and rules of thumb, which are exactly the
approaches this industry is trying to improve. It is this
approach that significantly diminishes the experience of
non-typical users, and potentially introduces risk while
not actually improving the design.
Furthermore, when gathering the relevant data for
these models, it is necessary to be aware of various ethical considerations. While not impossible, performing
experiments with vulnerable populations (children, for
example) raises a whole host of questions that need to
be answered in front of an ethics committee before carrying out these experiments. Without this data, the models will not be able to recreate scenarios accurately, but
in order to obtain it, the researcher must tread carefully.

Potential Benefits: Usability
Cities have not historically been designed to accommodate users with differing mobility issues, and they

IEEE TECHNOLOGY AND SOCIETY MAGAZINE

IEEE Technology and Society Magazine - December 2019

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - December 2019