Introduction

Over the course of several years, between 2020 and 2022, the authors of this book collected images of wild mammals in the suburban woodlands of Winnipeg, a city in the Canadian Prairies located midway between the two coasts. This was a COVID project that arose out of a desire to observe what we could of the behaviour and activities of animals in our suburban neighborhood and nearby locales. All the locations were easily accessible by bicycle and the upkeep of trail cameras throughout the year gave us an excuse to get out of our houses, meet our neighbours, and explore the many isolated and contiguous woodlands of Winnipeg’s southern suburbs. We purchased ten trail cameras and identified locations in pocket parks and along the riparian corridor of the Red River, a major continental waterway that flows north through Minnesota, North Dakota and Southern Manitoba before eventually emptying into Lake Winnipeg.

The participants in this project are professors and students in the Faculty of Architecture at University of Manitoba with an interest in the interaction between people and non-human animals in cities, and in simple methods for designers to collect information about the cities and neighborhoods in which they work and live. We had read about design projects in which the needs of wild animals were studied as a means of informing the design of the built environment, and we collected examples of design informed by animal lives. As designers living in a time of ecosystem destruction and mass extinction, we wanted to find ways to design not only with human needs in mind but also the life patterns and desires of non-human animals. Rather than trying to imagine what these patterns and desires might be, we set out to observe animals as best we could, gathering data to inform an evidence-based approach to design for humans and non-human animals.

We sought to identify methods that are inexpensive to implement and accessible to a broad audience. We found animal images to be an abundant source of information that yield different types of information to different audiences. For some members of our team the images told a story of animal lives that could best be teased out through a manual process of scrolling through images on the computer, one by one. Others in the team were interested in using machine learning to identify animals in images and automate the process of extracting quantitative data from the images.

The quantitative approach is the focus of this book and the associated resources, which introduce computational methods for automating the extraction of relevant information from large numbers of images. The methods introduced here could easily be adapted to other domains, and we have pointed at various points to these opportunities. Our focus throughout has been on the interests of designers in the built environment including landscape architects, architects, and urban designers, who can benefit in many ways from understanding animal lives in the places where they work and propose new built projects. This project has been enlightening for us about the many connections that we have with non-human animals, and the many ways in which our lives intersect or exist in parallel, occupying the same or adjacent spaces in the suburbs.

The topics covered in this book are a record of our team’s investigations over several years, investigations that were driven by personal and professional interest and that are certainly incomplete. While our work only partially and superficially explores human-animal interactions and emerging image pattern recognition technologies, we hope this organized collection of materials will encourage others to gather data in their own neighborhoods. We would be grateful for any reports from readers about errors and opportunities for improvement in this book and the accompanying resources.