Animals  have been involved in machine interactions for many decades. Skinner's famous operant conditioning chamber, used in behavioral experiments since the early 1930s , provided output devices, such as lights or sounds, and input devices, such as levers or buttons, and would dispense food or water, if, for example, a rat or a pigeon completed a given sequence of tasks correctly. These systems have gradually evolved into sophisticated computerized environments affording complex interactivity. Other interaction systems, such as computer games currently employed in more advanced primate cognition studies, provide, for example, on-screen animations that can be controlled via joystick .
Within agricultural engineering, interactive computing devices have also been developed, for example, to optimize milk production in the farming industry, with the introduction of the first automatic milking systems in dairy farms emerging in the early 1990s . These systems have rapidly developed into cutting-edge applications of pervasive and ubiquitous computing technology, enabling cows to independently engage in voluntary milking and express intelligent and social behavior never previously observed in constraining farming environments .
Examples of a different kind of interaction are provided by tracking and telemetric sensor devices, which have been used in conservation studies since the early 1970s and which have now become commonplace. For example, radio collars allowed researchers to uncover the elusive behavior and territorial needs of snow leopards for the first time , and satellite collars enabled conservation efforts to start mapping the movements of elephants . Tracking devices have also been introduced to the pet market, while various telemetric technologies are used in laboratory settings to monitor, for example, dogs' physiological parameters during pre-clinical trials .
In short, animal-computer interactions have a long history and can be found in many areas in which human activity involves other species.
In spite of its history, the study of the interactions between animals and computing technology has never entered mainstream computer science, and the animal perspective has seldom informed the design of animal computing applications, whose development has so far been driven by academic disciplines other than computer science or by other industrial sectors. The design of these technologies remains fundamentally human centered, and the study of how they are adopted by or affect their users remains fundamentally outside the remit of usercomputer interaction research.
The negative effects of this lack of animal perspective become obvious when, for example, the behavior and welfare of seals fitted with bio-logging tags and satellite transmitters are significantly affected and data gathered during costly conservation studies risks invalidation , or when cows who do not engage with milking systems are culled  and farmers suffer capital losses. But risk mitigation aside, what about the things we could gain from a shift in perspective? What would it allow us to learn about and achieve with interactive technology? How would it influence our reflection on usability, adaptation, appropriation, methodology, and ethics, to name but a few aspects?
Studies in interspecies computer interaction have started making appearances at HCI venues [10,11,1213], but the remarkably marginal position this research still occupies in the HCI community and its research agenda is an indicator that its significance has not yet been recognized. For some reason, animal-computer interaction (ACI) is, quite literally, the elephant in the room of user-computer interaction research. The time has come to acknowledge the elephant, to start talking about ACI as a discipline in its own right, and to start working toward its systematic development.
Advances in our understanding of animal and comparative cognition, as well as those in computing technology, make the development of ACI as a discipline both possible and timely, while pressing environmental, economic, and cultural changes make it desirable.
From long-held training experiences, we know that several species can use interactive devices of one kind or another, sometimes appropriating them in interesting and unexpected ways. More generally, though, we now know that many species have sensory faculties superior to ours , possess sophisticated cognitive abilities, engage in advanced problem solving, use purpose-built tools for complex tasks , communicate through articulated languages, experience a range of emotions, form complex social relationships, make moral judgements , and hand down cultures through generations . This has progressively made us more aware of the similarities between humans and other species, more appreciative of other species, and more attentive toward the significance of our relationships with them and the fragile environment we all share .
At the same time, the interaction modes afforded by computing technology have expanded well beyond those provided by keyboard and mouse. Tangible, embodied, and proxemic interactions, for example, have brought physicality back into computing by engaging the whole body through contact and movement. Sensor technology has become more agile, robust, and sensitive, better able to read the changes coming from within and around us. In general, developments in pervasive, ubiquitous, and ambient computing are enabling technology to adapt to our spontaneous behaviors and to the contexts that these continuously produce and modify. Not only do these advances make computing technology more accessible to humans but they also make it far more accessible to other species.
ACI aims to understand the interaction between animals and computing technology within the contexts in which animals habitually live, are active, and socialize with members of the same or other species, including humans. Contexts, activities, and relationships will differ considerably between species, and between wild, domestic, working, farm, or laboratory animals. In each particular case, the interplay between animal, technology, and contextual elements is of interest to the ACI researcher.
ACI aims to influence the development of interactive technology to:
- improve animals' life expectancy and quality by facilitating the fulfillment of their physiological and psychological needs; technology that encourages healthy feeding habits in domestic animals or allows them to modify their housing conditions at leisure might be consistent with this aim;
- support animals in the legal functions in which they are involved by minimizing any negative effects and maximizing any positive effects of those functions on the animals' life expectancy and quality; technology that gives farm animals control over the processes in which they are involved, produces no side effects on the animals involved in conservation studies, or helps working animals communicate with their assisted humans might be consistent with this aim; and
- foster the relationship between humans and animals by enabling communication and promoting understanding between them; technology that allows companion animals to play entertaining games with their guardians or enables guardians to understand and respond to the emotions of their companion animals might be consistent with this aim.
ACI aims to develop a user-centered approach, informed by the best available knowledge of animals' needs and preferences, to the design of technology meant for animal use. It also appropriately regards humans and other species alike as legitimate stakeholders throughout all the phases of the development process.
ACI takes a non-speciesist  approach to research, and researchers have a responsibility to:
- acknowledge and respect the characteristics of all species participating in the research without discriminating against any of them;
- treat both human and nonhuman participants as individuals equally deserving of consideration, respect, and care according to their needs;
- choose to work with a species only if the intent is to advance knowledge or develop technology that is beneficial or otherwise relevant to that particular species;
- protect both human and nonhuman participants from physiological or psychological harm at all times by employing research methods that are noninvasive, non-oppressive, and non-depriving;
- afford both human and nonhuman participants the possibility to withdraw from the interaction at any time, either temporarily or permanently; and
- obtain informed consent to the involvement of both human and animal participants, either from the participants themselves (for example, for adult humans) or from those who are legally responsible for them (for animals).
The development of ACI as a discipline could have many benefits for both animals and humans. For example, it could have important effects on our interspecies relationships by informing the design of technology that enables the animals we live and sometimes work with to effectively communicate with us, increase their participation in our interactions, and constructively influence our environments. These developments could give us a better understanding of those we share our lives with and help us build safer, richer, longer, and more productive relationships with them.
ACI could also lead to further insights into animal cognitionfor example, by informing the design of interactive technology for behavioral studies that affords optimal usability and creative appropriation for the animals. Or it could support conservation effortsfor example, by informing the design of monitoring devices that minimize the impact on the animals while maximizing the quality and reliability of the data gathered through them.
Moreover, ACI could improve the economic and ethical sustainability of food productionfor example, by informing the design of technology that affords farm animals more freedom and autonomy, enabling them to live less unnatural lives, reducing their stress levels and susceptibility to illness without recourse to drugs, increasing their productivity, and improving the quality of their produce.
Finally, ACI could expand the horizon of user-computer interaction research by pushing our imagination beyond the boundaries of human-computer interaction. For example, it could help us discover new ways of eliciting requirements from those who cannot communicate with us through natural language or abstract concepts. It could help us explore new modes of interaction for those who do not have hands, cannot decipher the patterns emitted by a screen, or have limited attention spans. Or it could help us find new ways of understanding and evaluating the impact of technology on individuals and social groupsperhaps shedding new light on issues such as identity, privacy, or trust, and contributing to our understanding of what it means to be human and who we are in relation to other species.
Of course, whether ACI can yield the benefits outlined here depends on our ability to tackle some challenging questions. For example, how do we elicit requirements from a nonhuman participant? How do we involve them in the design process? How do we evaluate the technology we develop for them? How do we investigate the interplay between nonhuman participants, technology, and contextual factors? In other words, how are we going to develop a user-centered design process for animals? Here is a possible roadmap:
- First, we could look at what has been done in other areas, what knowledge about animal behavior and psychology is available, and what data has already been collected about ACI. We could look at how all that maps onto what we know about user-computer interactions and how it might contribute to ACI as a discipline and design practice.
- Second, we could form collaborations with researchers from disciplines such as ethology, behavioral medicine, animal psychology, and veterinary, agricultural, and environmental engineering to help us with this mapping effort. Similarly, the expertise and experience of professionals and practitioners who work with animals in environments where ACI take place would be important.
- Third, we could study in-the-wild cases of whatever technology is already in use or might be developed in order to understand those domains and contexts, their users, and their stakeholders, so that we can begin to develop or adapt relevant ACI concepts and models.
- Fourth, we could look at human-centered interaction design protocols and methods to assess which ones may or may not be relevant to an animal-centered design process, which might be adapted, which might be borrowed from other disciplines, and which might need to be developed from scratch.
- Fifth, we could start adapting, developing, and integrating ACI design protocols and methodsfor example, for requirements elicitation, participatory design, and contextual evaluation, in a loop between empirical work and theoretical reflection.
- Sixth, we could start developing theoretical models of ACI, which would then drive further research. These would take into account pre-ACI research on animals and would be informed by ACI empirical research with animals.
Because of the questions it raises and the challenges it poses, ACI is arguably the next frontier in the study and development of interactive technology. Those who are keen on joining in the exploration of this new territory are warmly invited to sign the ACI Manifesto and join our animal-computer interaction group at: http://www.open.ac.uk/blogs/ACI/
I am indebted to Yvonne Rogers, Bashar Nuseibeh, Marian Petre, Anne De Roeck, Hugh Robinson, Janet van der Linden, Richard Power, Shailey Minocha, Sandra Williams, Daniel Mills, Shaun Lawson, Helen Sharp and Simon Buckingham Shum for their constructive criticism and support. This work was supported by the EPSRC grant EP/F024037/1.
2. Operant conditioning chamber; http://en.wikipedia.org/wiki/Operant_conditioning_chamber
3. Gill, V. Monkeys "display self-doubt" like humans. BBC Earth News, Feb. 21, 2011; news.bbc.co.uk/earth/hi/earth_news/new-sid_9401000/9401945.stm
5. Brennan, Z. Ooh-aaar, cows do milk themselves, down on the robotic farm. The Sunday Times, Oct. 23, 2005; http://www.timesonline.co.uk/tol/news/uk/article581764.ece
8. Emka Technologies. Telemetry in primates; http://www.emka.fr/telemetry-in-primates-94.html
10. Lee, S.P., Cheok, A.D., and James, T.K.S. A mobile pet wearable computer and mixed reality system for human-poultry interaction through the Internet. Personal and Ubiquitous Computing 10 (2006), 301317.
12. Noz, F. and An, J. Cat cat revolution: An interspecies gaming experience. Proc. of the 29th International Conference, Human Factors in Computing Systems (Vancouver, BC, CA, May 712). ACM, New York, 2011.
13. Weilenmann, A. and Juhlin, O. Understanding people and animals: The use of a positioning system in ordinary human canine interaction. Proc. of the 29th International Conference, Human Factors in Computing Systems (Vancouver, BC, CA, May 712). ACM, New York, 2011.
14. Willis, C.M., Church, S.M., Guest, C.M., Cook, W.A., McCarthy, N., Bransbury, A.J., Church, M. R.T., and Church, J.C.T. Olfactory detection of human bladder cancer by dogs: Proof of principle study. British Medical Journal 329, (2004).
Clara Mancini is a research fellow in the department of computing at the Open University, Milton Keyes, UK, where she investigates how computing technology affords, mediates, and shapes new forms of communication and sociality. She is interested in expanding the boundaries of interaction design.
©2011 ACM 1072-5220/11/0700 $10.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2011 ACM, Inc.