Why we need to design technology for and with animals

Josie: Good evening and thank you for joining us tonight for another lecture in our inaugural lecture series. I'm Josie Fraser. I'm Deputy Vice-Chancellor at The Open University, and I'm proud and privileged to get to host tonight's lecture. Our inaugural lectures are a series that showcase our research, our teaching and our knowledge exchange as a university. Today we're going to hear from Clara Mancini, who is our Professor of Animal-Computer Interaction in the OU’s Faculty of Science, Technology, Engineering and Mathematics. Clara's going to talk to us about her research journey in the field of Animal-Computer Interaction, shared with colleagues from across the world, she has developed the art and science of designing technology for and with animals.

Before we begin a little bit of housekeeping. If you're with us today in the Lecture Theatre in Milton Keynes you can join us for refreshments downstairs. Whether you're in the Lecture Theatre or online with us today, Clara's lecture will be followed by a Q & A session with a special guest for those of you that are watching or in the room. If you're in our audience and you're using X, formerly known as Twitter, please feel free to post using our hashtag, #OUtalks and tag @OpenUniversity, and we'll get lots of interaction going online around the world today. For those of you that are joining us via YouTube, please use the email address on the slides and you'll be able to send us your comments or questions so that we can address those with Clara in the Q & A after her talk.

Now it's time to introduce Professor Clara Mancini. Clara is the Professor of Animal-Computer Interaction and the founding Head of the Open University's Animal-Computer Interaction Laboratory. Her work investigates the interaction between animals and technology, and the nexus between technology, animal wellbeing and justice, and thinks carefully about human-animal relations. Her research has spanned theory, methodology, practice and ethics of designing animal-centred interactive systems, not just for animals, but also with animals. Clara wants to inform a more equitable and inclusive multispecies society. Her animal-computer interaction projects that she’s supervised and led have benefited both managed and wild animals, and her collaborations with industry and non-governmental organisations have resulted in some very practical innovations. She has published extensively in the leading peer-reviewed journals across a wide range of disciplines, from computing and interaction design to animal-computer interaction, animal behaviour and animal welfare, and she's lectured and refereed on these topics worldwide. Her work has reached broad audiences through public lectures like this one today, and also through exhibitions, including demonstrating some of her work for Her Majesty Queen Elizabeth II, and she's been featured in national and international media, including print, radio and television. So I think we've all got something to really look forward to today, and it gives me great pleasure to introduce Professor Clara Mancini.

Clara: Thank you so much Josie and thank you all for joining us this evening. Before I begin, the work that I'm going to present is the result of national and international collaborations with colleagues from academic institutions, charitable organisations and industrial partners. I am profoundly grateful to these collaborators for sharing with me in this challenging, exciting and important journey, and for making me a better researcher in the process. In the interest of time, I can only give you an overview of selected projects and collaborators on these are acknowledged along the way, but my heartfelt gratitude also goes to the many colleagues who make up the wonderful animal-computer interaction community.

In a primordial landscape great apes survive by forging alongside other herbivores, by dodging the attacks of carnivores, and by competing with rival troops for access to water. One fateful dawn, the apes awakened to the sight of a monolith, its geometric dark form stark against the rugged deserted landscape, a rising sun and crescent moon perfectly aligned above it. The alien intelligence that the object embodies changes the way the apes look at the world. So when next rummaging for food among scattered bones, one of them suddenly realises that a femur can be used as a tool to kill the animals alongside whom he had been foraging, and who will thus become his prey. Soon after, all members of the apes’ troop brandish the same kind of weapon which allows them to defeat a rival troop by killing one of its members and scaring away the others, and thus seize control of the local water source. From then on, the dawn of man is set onto a path of technological development that will lead him millions of years later to colonise space.

The dawn of men, the opening sequence of Kubrick's film ‘2001: A Space Odyssey’, famously provides a poignant imagining of the moment when humans discover technology. The juxtaposition of the sequence's closing shot, a bone thrown in the air by one of the apes, with the next sequence's opening shot, the elongated shape of a military satellite orbiting space, convey the transformative function that technological innovation has had in human evolution, as well as its inherent association with violence.

Indeed, much of human technology has advanced human interests by making it increasingly easier for us to outcompete and exploit other animals. With the advent of computing and with developments such as the Internet of Things and Artificial Intelligence, animals are being increasingly impacted by technologies whose design is still fundamentally human-centred.

In contrast the growing field of Animal-Computer Interaction, or ACI for short, focuses on the nexus between animals and computing technology from an animal-centred perspective.

In particular, The Open University's Animal-Computer Interaction Research Programme was set up in 2011 on a mission to “advance the art and science of designing animal-centred interactive systems for a multispecies society”.

Its aims are studying different kinds of interactions between animals and technology within their habitable context, designing computing technology that improves animals’ welfare, supports animals in their activities, and fosters intra and interspecies relationships, and developing animal-centred approaches to design technology for and with animals as legitimate stakeholders and design contributors. With this in mind, it is important to understand how even the best intentioned users of technology may impact animals and how to minimise these impacts.

For example, animals are often fitted with devices that track their movements, behaviours and vital signs for conservation, research and husbandry purposes. However, animal wearers, the primary stakeholders, have no control over their use. Devices can interfere with their body, for example, injuring or trapping them and activities, for example, slowing them down or disrupting their camouflage, thus impacting their welfare and the quality of recorded data. Existing guidelines for designing these devices still tend to be rather coarse and ad hoc.  

This project investigated how to minimise the impact of wearable animal trackers by improving what we call ‘wearability’. We adapted values and principles informing the design of human wearables to develop a framework that could support a wearer-centred approach to the design of wearables for animals. The framework is underpinned by the fundamental values of non-deleteriousness and minimisation of effects, and by three design principles stating that the presence of tracking devices on the animal's body should ideally be imperceptible to them, or at least unobtrusive, or at least cognitively acceptable to them. All stakeholders should be considered, primarily the wearers, but also what we call their ‘significant others’. For example, social relations such as mates and offspring, and asocial relations such as predators and prey, and, of course, human users too. Whereas biological characteristics, their activities and their environment should all be considered and all possible design variables, along with technical variables, should be exploited to maximise wearability. The example on the right suggests that when designing a tracker for red foxes, the device should not emit background noise that is within the hearing range of the fox, but also of her potential prey and predators. This is just one aspect to consider.

As a case study, Patrizia's behavioural observation of cats wearing different commercial GPS trackers showed poor wearability, even though these were marketed as cat friendly. In contrast, her behavioural observation of the same cats wearing a tracker that she had redesigned using the same electronic components but applying the framework, showed significantly better wearability.

Collecting animal data in non-intrusive ways is particularly important when animals are asked to perform tasks that require concentration and precision. For example, bio-detection dogs are trained to recognise and indicate the presence of organic volatile compounds from diseased cells in biological samples such as urine or sweat. But the signalling conventions they are trained to use for human convenience, such as sitting down in front of a positive sample, do not enable them to communicate all that they know, because these conventions are binary signals that do not express nuances, such as a compound's concentration levels, and because they impose a ‘translation’ overhead on the dogs that lowers this signal's reliability.

With Medical Detection Dogs, the charity that pioneered bio-detection with dogs, we wanted to find a way of accessing the dog's knowledge about the samples they sniff without them having to translate. We wanted to centre the practice of bio-detection on the dogs by shifting the translation burden onto the technology.

Hypothesising that dogs’ spontaneous response to the olfactory stimulus might be proportional to the stimulus presence and intensity, we designed a device that used pressure sensors to record the dog's interaction with the samples and capture their stimulus response.

Early tests with solutions of isoamyl acetate, that is banana essence, showed some consistency in stimulus response between different dogs, with an interaction pattern emerging when the dogs encountered positive samples. It seemed that there might be proportional variation in stimulus response depending on sample concentration. This suggests that the data may lend itself to machine learning approaches, and a preliminary application of neural networks to classify the dogs’ responses has already yielded promising results.

Medical Detection Dogs are currently using the latest version of the technology in their colorectal cancer detection project. The new apparatus captures both pressure and proximity data which we are looking forward to analysing soon.

While here the aim is to allow dogs to focus on their work without taking any notice of the technology. In some cases dogs have to engage with technology to negotiate their way around their living and working environment. For example, mobility assistance dogs that are trained to interact with various technologies on humans’ behalf have to operate in a human-centred environment that, from a dog's perspective, contravenes the most fundamental interaction design principles. This results in poor canine usability and user experience and impacts not only on the dogs’ training times and performance, but also on their welfare.

With Dogs for Good, a leading charity in the sector, we wanted to rethink the way in which ambient interfaces are designed to make them more accessible for dogs. We wanted to develop a portable kit of controls that could be retrofitted in the dog's environment and living and working environment to make it easier for them to work within it.

We applied established interaction design principles based on dogs’ characteristics, including their sensory abilities, such as their dichromatic vision, their cognitive abilities, such as the way they categorise objects, and their physical abilities, such as their snout’s dexterity versus their limb strength. We also needed to take into account the context in which these devices need to be used, including the dog's activities, for example, opening doors, switching lights, calling elevators and their environment, that is their domestic and urban settings.

Our research shows that input devices designed according to this criteria made dogs’ tasks much easier. For example, Luisa's comparison of our early controls against standard issue access controls for opening a motorised door showed improved usability based on parameters such as, overall time to open the door, number of contact attempts made, and attempts necessary to activate the control. Here is a short video of my dogs using the same controls after only a couple of short training sessions.

[video]

Clara: The first is Zena, the second is Cara.

So Luisa further developed the design of our next generation of wireless cylindrical controls testing different sizes, as well as push pads with concave and convex profiles. She found that a control with a diameter 120mm and convex push pads performs best for dogs of different sizes. In particular, although the concave pads guided the dogs now towards the centre of the control, inviting a more precise interaction, the convex pads enable the dogs to activate the control from any angle. This provided more flexibility regarding their placement in the physical environment, which is critical for usability in small spaces.

Following further improvements, here is the Dogosophy button, our final wireless prototype which can be used to activate a range of domestic appliances, and which we are currently in the process of commercialising with Ravencourt.

Would anybody like to say what this is?

Speaker: The sign of Stereotyping.

Clara: That's right. So while, as we have seen, sometimes the environment in which animals operate can be too complex. Other times the settings in which they find themselves lack the complexity and stimulation of the environment in which they evolved. Many wild animals who live in captivity develop stereotypes such as endlessly walking in an infinite configuration. The problem is particularly evident in some large carnivores, and tigers are a case in point.

In the wild tigers have extensive home ranges. They spend their time patrolling their territory, foraging, hunting, mating, all the while reading the myriad signs of the land. In contrast, in captivity they have limited space, and everything is provided for them on a schedule, so they have no need, no opportunity to engage in activity that in the wild would be essential and involve meaningful movement. They still cover long distances by pacing in their enclosures, but they are not going anywhere and do not engage their cognitive capacities as they could.

Research conducted by Care for the Rare found that meaningful movement is a major welfare priority for tigers, which led to Jake's suggestion to design a whole interactive environment to provide tigers rescued by the international animal charity Four Paws with opportunities for meaningful movement.

TERA, the sanctuary where the tigers reside, is creating a new complex of interconnected outdoor enclosures. Three of these are connected via a large building where our interactive system is soon to be installed.

The building will host an interactive maze consisting of three concentric irregular rings and an elevated central island. The rings will be connected by two diametrically opposed gate systems, which will open and close to reconfigure the space and provide multiple routes to and from five different locations, three outdoor enclosures, one indoor den and the elevation of the central island. As the tigers travel through the maze, olfactory, acoustic and visual signals associated with different locations will guide them around the perimeters and through the gate. At specific points the gate’s configuration will present the tigers with route options, giving them the opportunity to make choices to reach specific locations and the resources within them. Infrared sensors all around the rings’ perimeters will allow the system to identify the tiger's location and trajectory, and to dynamically adapt gate configurations and signals accordingly for a seamless journey. To our knowledge this is the most complex interactive environment ever developed for animals, and we are looking forward to installing it in early September when we will be able to begin working with the tigers themselves to adapt the functionalities of the system to their learning and experience requirements.

Indeed, a tenet of ACI is that whenever possible designing for animals is best achieved by designing with them. In this regard, the ethical and methodological perspective with which we approach the process is key. However, legal frameworks currently regulating the involvement of animals in research are fundamentally process-centred, rather than animal-centred.

To support animal-centred research and design we needed a different ethical approach, So above and beyond existing regulations, our ethics framework is grounded in four key principles. Relevance, which requires working only with animals for whom the research is directly relevant and beneficial. Impartiality, which requires protecting all participants in virtue of their role rather than their characteristics, such as age, sex, species or assumed levels of sentience. Welfare, which requires us ensuring all participants’ physical and psychological wellbeing at all times. Consent, which needs to be obtained from all participants, including animals.

Importantly for us, animals’ consent takes two complementary and equally necessary forms. Mediated consent obtained from human guardians who know the animal well and understand the implications of their involvement, who have their best interests at heart and the legal authority to decide on their behalf, and contingent consent provided by the animal participants themselves, who must be allowed to continually assess the situation and choose what to do, whether and how to engage, while appropriate monitoring of any manifestations of dissent should enable guardians to respond accordingly. In a nutshell, doing animal-centre research and design implies respecting participants’ autonomy and integrity at all times. Working in a context that is habitable for the animals being unobtrusive and undisruptive of their habitable activities, giving animals control and space for expression, and using only positive forms of interaction.

This approach is applied within an interaction design process that is quintessentially iterative, and during which design requirements are elicited using various methods. Design concepts are produced. Selected designs are prototyped and then evaluated to identify further requirements. This is an open-ended process of incremental orientation towards a perfect final outcome that may never be achieved, but that can be gradually approximated over several iterative cycles. It is in this context that animals’ ethical involvement becomes participation as it informs the design process.

As an example, we explore questions regarding animal participation by teaching my dog, Zena, to interact with the early versions of our dog-friendly controls to open a small door and switch on a light. The coming short videos offer an example of how we take input and direction from our participants.

[Video]

So here we are already making progress with the training which employs the luring method. To help Zena associate her touching of the control with the reward, I place a treat on top of it. So whenever she goes for the treat, she simultaneously triggers the control, which soon establishes a consistent association between the two events. At the same time this creates a strong bias towards snout interaction. To make the opening of the door relevant for Zena you see rewards are through the door when this opens, but once she has learnt I only reward her after she has triggered the control.

Now here I'm trying to persuade Zena to touch the larger yellow light switch control that is above the small door, which is a much more demanding task due to the control's height. She has done it a few times already, but now she refuses to continue, sitting down and looking at me in a begging behaviour instead.

Here she still refuses to engage with the light switch control. I could offer a higher value reward to persuade her, but keeping the reward constant tells us that there is an effort threshold that we should not pass. Instead, I ask her to touch the lower yellow door opening control. Despite the bias towards snout interaction that I had previously created she chooses to use her paw. It is clear that she understands what the task entails, but she elects to do it in her own way. She interprets the task and the interaction patterns required to complete it, and by rewarding her I accept her interpretation. This has implications for design that will inform the next iteration cycle. For example, that the control needs to be located much lower obviously, that the pad needs to be sensitive enough to respond to snout interaction which can be very light, and robust enough to withstand paw interaction which can be very heavy, and that the pad surface needs to be comfortable enough for a dog's delicate nose to touch, whilst also withstanding claw abrasion.

For another example, in this project we were aiming to design an alarm to enable medical alert dogs trained by Medical Detection Dogs to call for help in a crisis. Although these dogs consistently alert their assisted humans to upcoming critical medical episodes, the humans are not always able to respond in time and may become incapacitated which leaves the dogs powerless and in distress. The dogs needed to interact with the alarm as autonomous users, so the interaction needed to be safe, efficient, easy to learn and to remember.

This particular design concept comprised an electronic module that would trigger a call to a phoneline and a biteable tag for the dog to use as an input device connected to the electronic module by a trigger. But what should the tag be like? What should the size, shape, texture, material be? What kind of trigger should connect it to the electronic module? Should the tag detach when pulled, or should it stay in place? How should the system give the dog feedback to let them know that they have successfully completed the task and raised the alarm?

To answer these questions, Charlotte used what we call quick and dirty prototyping to engage in physical conversations with canine users. She formulated a modular template for a canine alarm consisting of a tag for the dogs to use as an input device, a block of wood in lieu of the alarm, an electronic module, and the trigger connecting the two. In this way, she could quickly present the canine participants with successive combinations of different tags and different triggers.

This dialogue suggested that the dogs preferred a simple and tidy set-up not to confuse them. A detachable tag that worked as a form of feedback letting them know that they had completed the task, and a customisable tag and trigger resistance depending on the dog's age, training stage and personality. Here is an example of a canine user in action with a customised tag.

Now working with animals is even more important when aiming to design creative and enriching forms of interaction for species whose physical, sensory and cognitive characteristics are very different from those of designers and with whom we have no familiarity. Especially when designing interactive systems for playful enrichment, the interaction needs to be intrinsically rewarding regardless of the outcome. In this regard the sensory experience with input devices plays a critical role.

In Fiona's work engaging directly with elephants over an extended period of time, first observing them in their environment and then offering them a wide range of prototypes over multiple iterations, highlighted the importance of considering the animal's aesthetic experience, which was instrumental to inform the design of appealing interactions. Here you see only a few examples of many prototypes designed to enable elephants to control sound output, including wood and textile fabric buttons, knitted buttons and copper radio buttons.

Here is an elephant ‘jukebox’ which uses copper radio buttons to trigger different pre-recorded sounds, and which was installed in the indoor habitat of two elephants residing at Noah’s Ark Zoo.

So far we have looked at technologies broadly designed to support animals with a focus on their needs. But what about situations in which the stakes for animals are higher? Is it possible to design animal-centred technology when designers operate in socio-economic systems that are not animal-centred? Addressing this question requires dealing with the wider political implications of animal-centred design. This means being aware that design is an intrinsically political activity in which decisions are made that change the world and impact different groups of stakeholders. This inevitably raises the issue of justice in general and on multispecies justice in particular, which we explored in this project.

For this we refer to Martha Nussbaum's political philosophy and to her capabilities approach, according to which animals are agents capable of dignified existence with corresponding needs for flourishing and related goals they actively pursue to achieve capabilities to which they are morally entitled. The approach does not demand interspecies equality or the abolition of all instrumental relations with animals, but it does demand that the fulfilment of animals’ capabilities to at least an adequate threshold as relevant to their species. It also supports respectful practices that are compatible with animals’ capabilities and that develop their excellencies and promote their flourishing.

The capabilities approach to multispecies justice provides a measure of animal-centredness that could inform wider considerations when designing animal technology or when studying animals' interactions with technology. Here are some examples of questions that could help researchers assess the extent to which their work is animal-centred and thus conducive to multispecies justice. Additionally, among others, we proposed three design strategies to promote animal-centred design.

Disruption involves questioning systems, norms and practices that go against animals' basic capabilities and that designers would like to address. For example, Natalia and I questioned whether smart dairy farms of the future might breach animals’ entitlement to privacy and non-exploitation. Entitlements which can be successfully argued for by using already existing jurisprudence.

Reconfiguration involves redefining design frameworks whose current human-centric definition legitimises the dismissal of animals’ role in the design process, or justifies design outcomes that discriminate against them. For example, UCNI proposed a multispecies framework for participatory design to account for animals’ agency and interest in design processes and outcomes.

Pollination involves seeking opportunities to enable animal-centred values to influence systems and practices which involve animals but which are not animal-centred, to have such systems evolve towards animal justice. For example, Eleonora and I have been exploring the extensibility of ACI’s animal-centred research ethics to all animal research. We have developed an ethical review framework that integrates the principles of relevance, impartiality, welfare and consent for the iterative refinement of any research activities involving animals.

One example of how these strategies could be applied is by reinterpreting governance ethical frameworks from a multispecies perspective to foreground impacts of emerging technologies on animals and raise animal-centred ethical concerns. For instance, what if we assessed smart farming systems against the principles underpinning the European Commission’s Ethics Guidelines for Trustworthy Artificial Intelligence. The guidelines specify that designers of AI systems should, “pay particular attention to those who have historically been disadvantaged or are at risk of exclusion, and to situations which are characterised by asymmetries of power or information.” A statement that perfectly describes the conditions of farmed animals.

Global governance bodies are increasingly recognising the interdependence of human, animal and environmental well-being and the inability of human-centric approaches to deliver sustainability. Arguably an animal-centred perspective on the design, development and deployment of computing technologies could support the pursuit of global sustainability agendas. For example, much animal-centred design is highly relevant to several of the United Nations’ Sustainable Development Goals. Consider the goal of good health and well-being. Animal-centred technology could foster beneficial human-animal relations. It could support animals as well as humans or improve animal welfare, reducing the risk of disease. Consider life below water and life on land. Animal-centred technology could engage people in conservation activities, enhance conservation efforts and reduce human-wildlife conflict. Or consider sustainable cities and communities. Animal-centred technology could foster beneficial human-nature connections within urban environments and facilitate social inclusion by involving marginalised groups in urban-wildlife custodianship.

In this regard, I would like to end with a vision which brings together some of the points discussed so far, and which resulted from a collaboration exploring the role of technology such as Artificial Intelligence in the development of Multispecies Smart Sustainable Cities as a matter of justice. Now conceptually smart sustainable cities are expected to integrate advanced information and communications technology with sustainable urban infrastructures. Digital sensing technologies, cloud computing infrastructures, middleware architectures and wireless communication networks would be used to monitor and analyse urban phenomena and to plan and manage access to resources and services in order to meet the environmental, economic, social and cultural needs of their citizens with equity and inclusivity.

But urbanisation is increasingly encroaching on animals’ habitats, and animals are moving into cities, which have an increasing role to play in biodiversity conservation. Thus, in order to be sustainable, smart cities need to acknowledge animal dwellers citizenship and meet their needs too. In other words, we argue that smart sustainable cities should enable animal dwellers to equitably and inclusively perceive, make sense of and interact effectively with their urban surroundings as they pursue biologically relevant goals, including environmental goals, such as building, inhabiting and traversing structures. Economic goals such as acquiring and managing resources. Social goals, such as forming and maintaining social groups. Cultural goals such as gaining and transmitting knowledge and skills.

But how to determine what goals different species value and when these are adequately met How to identify species to be considered, and how to decide which species’ needs to prioritise? How to reconcile conflicting needs among species and compensate for lost opportunities? How to unambiguously communicate relevant possibilities to different species?

This is where smart technology could really help. Starting small, maybe do one species or a few in a location or two and iteratively expanding to encompass entire urban ecosystems using multimodal sensors to collect data on animals’ presence, activity and existing possibilities. Leveraging various capabilities of Artificial Intelligence to process and interpret data for identifying needs and trade-offs. To inform design solutions that mediate different species’ needs and to select solutions that maximise benefit where most needed and to communicate relevant possibilities to different species in a species appropriate way. Imagine an ecologically and ethically responsible AI that redistributes control and benefits across all city dwellers.

At the end of 2001 after the monoliths last appearance, Kubrick leaves us with a vision of a giant foetus floating in space bathed in cosmic light, eyes wide open, seemingly representing the rebirth of a new superior humanity and the promise of a renewed future. However, our reality is that no alien intelligence is likely to transform us anytime soon. The one we need to embark on with great urgency is not a Space Odyssey, it's an Earth Odyssey. It is up to us to use the intellectual and emotional capacities with which millions of years of evolution alongside our fellow earthlings has endowed us to address the existential challenges that face our planet, so that it can sustain all earthly life into the future. While technology can certainly play an important role in the pursuit of a sustainable future, this will arguably depend on whether we will continue to use it to exploit and out-compete other creatures, further alienating ourselves from the natural world and from the multispecies relationships that sustain us, or whether we will be able to leverage it to reduce our separation from nature and live in more harmonious relations with the beings with whom we share our common home. For the latter to happen, we need to start thinking outside the human box and stop prioritising human interests to everyone else's detriment. To this end, ACI’s animal-centred perspective could not only yield design outcomes and processes that improve animals’ lives and human-animal relations. But arguably also, and more importantly, it could produce ways of thinking and practices that are more attentive, sensitive and empathetic towards other beings helping us to meet the responsibility that arises from our ability to develop technology and from the power that technology gives us.

In brief, ACI provides us with an opportunity to reconfigure multispecies’ relations and redefine our place in the world for a more inclusive, equitable and sustainable future. Thank you.

Josie: Thank you very much, Clara, that was absolutely fascinating. So now it's time to hear from you in the Lecture Theatre, and those of you who are online with any questions. So Clara, hopefully you'll join me over there in the seating area.

Clara: Do you think it’s a little dark here?

Josie: Maybe.

Clara: Shall we ask Lucas to help us?

Josie: Well done Lucas I’m very impressed.

So we've got a microphone out for anyone who's got a question for Clara. If you can introduce who you are and where you're from that would be great. We've also got a couple of colleagues down here monitoring any questions or comments from our online audience, so don't be shy. Shall we start with one online.

Nicholas: This is one from the Executive Dean of STEM, Nicholas Braithwaite. Thank you for your stimulating talk Clara. There are two questions here, so I'm just going to read them both, because you may be able to just roll them into one. So either - it was exciting to hear that we have much to gain from parts of the animal kingdom with respect to sustainability, but are there examples of animals as well as humans living unsustainably from which we can both learn? Or you may want to consider - what has been the most surprising successful communication scheme, ie, channel of interaction that you've established in partnership with any species, though I am particularly interested in new ways to commune with my cocker spaniel.

Clara: Those are two very different questions I think. You say animals who live unsustainably we’ve come across. I think actually if there are animals who live unsustainably it is because of us. So what's happening, in nature animals don't get a chance to live unsustainably because the environment is really quite harsh. They have to work hard to get resources to live and everything is rather in balance. When we, for example, create in our city’s places where there's lots of rubbish that allows large communities of animals to scavenge to live off of those scraps. When there is a lot of them, it may be that they have an impact on other local species. Maybe they out compete them, or maybe they eat too any of them, or something like that. But that usually tends to be out of imbalances we have created. So the other question was about the communing with the cocker spaniel, so examples of communication. It depends on what you mean by communication and by communing. If you mean being able to exchange, to be able to interact and make sense of that interaction in a way that you can use in design, then that's based in several cases. Well, I worked more with dogs I suppose so that would be my prime example. But I think Nick can commune with his cocker spaniel by leveraging some associative learning practices, if it is at that level that he would like to commune with him.

Josie: I think we've got a couple of questions in the room.

Question: Hi Clara, thanks very much for the great talk. As you know I'm a big fan, and whenever you speak I always learn new things, so I was very excited to hear about the things you were describing on smart cities and species communication. I can see there's going to be conflicts where humans don't want the pigeons leaving messes on their cars, or urban foxes getting into things. Do you have any examples of work with smart cities where technology is being used to manage attention one way or the other and what do you do? Who do you prioritise in the end?

Clara: So far the actual practical work with smart cities is still pretty much human-centred. What I presented towards the end is something that would be nice to do in this way. But the basic idea is that we dissent the human so that the human doesn't get to decide anymore everything and that in a way Artificial Intelligence takes some of our control to redistribute it to other species, based on the global ecology of the city, or what everybody might need, and that's the point. Yes, people don't want pigeons to poop on their cars but do we realise that without pigeons the Second World War could have ended very differently? Are we aware of what we owe to other animals, the fact that we are not the most important, the only ones. What Martha Nussbaum effectively says in her political philosophy is that we need to accept that sometimes humans’ interest has to take a step back as to be more in the background, and that we have to allow the most basic capabilities of other animals to take priority, to the detriment of our non-basic capabilities. So if we have to clean up a little bit of pigeon poo, at the end of the day we need to live with that. That's the whole point. We have created a world in which we're not willing to, or very rarely, and only in cases where we really think a species is very cute and very useful, very rarely we are willing to give something up in order to allow others to enjoy essential capabilities, like staying alive for example. I think we could think of using Artificial Intelligence to try and work out these negotiations. I think it would be really interesting.

Speaker: Thanks. I'm looking forward to hearing more about it later.

Josie: So we've got another online question, and then there's another one in the room at the front here. Thank you.

Tamsin: This is from Tamsin Morris, one of our Associate Lecturers. A fabulously interesting lecture. Thank you. Have you engaged farmers or agribusiness in any of your work?

Clara: I am trying. I am finding that is a sector that feels quite challenged. I know they work on small profit margins and they are very focused on efficiencies. They are protective of their business, that's their livelihood. But I think there is scope to improve things for animals. I am actually making steps to try and work in that direction. It's not the easiest. Sometimes you have to work with who allows you to work with them, if that makes sense, and hopefully you learn from those projects, from those experiences, and hopefully then you can apply that in other contexts. So I'm really hoping that soon I'll be able to work with farm animals, I am actively pursuing that.

Arosha: Thanks very much. Clara, that was a real tour de force, and congratulations and well done on an amazing and fascinating talk. For the benefit of colleagues I'm Arosha Bandara, I'm a Professor of Software Engineering and the Associate Dean for Research in the Faculty of STEM. My question is a bit, well it's outside my area of expertise, so this is not very well thought out, but I'm curious, you really make a compelling case for this multispecies vision of the future, but it's one that takes political will to enact, and this seems to be a huge challenge. I just wondered whether you were aware of where to start driving that wedge in. Is there any glimmer of where there might be a door that could be pushed against in the political arena that would help us think beyond just a human-centred approach to the world?

Clara: It's really hard Arosha. It's really, really hard because we are holding on, you know. Sustainability, the narrative around sustainability is one. There is the One Health approach. There are the Sustainable Development Goals. There are areas. Very recently, like a couple of years ago, the United Nations for the very first time produced a resolution on the nexus between animal welfare and human wellbeing, and therefore the importance of considering animal welfare. Unfortunately, it's all still very much geared to, we care about sustainability because we think we want it for the future generations of humans. We care about the animal welfare because otherwise we get diseases. It's all very human-centred. But my view would be that any steps we can take forward in that direction are worth taking. Then once we are a bit more advanced maybe we can see things a little bit differently. The point is that we are all connected. Our world, the natural world that’s what we need. It’s a web of life. It's like tapestry. If we disrupt it we do damage to ourselves as well. So in a way it's realising that we are all part of the same thing. So it's not about throwing the human out of the window, but it's about repositioning the human so that the balance that is needed for everyone, including humans, can be established, reestablished, or whatever. We live in a world that is finite, and so there is this competition, there is disease, there are all these difficulties, but there are better balances than what we have currently, I think. I think in order to achieve that we need to stop putting ourselves on top of everything. We have to take whatever steps we can. I think the governance bodies are starting to wake up to the need for this.

Josie: We've got another one online, and then a question.

Question: So Tamsin comes in again and says, I see a link with rewilding. I wonder if Professor Clara agrees.

Clara: Yeah, it's interesting. Somebody just published a book about, it's titled ‘The Cull of the Wild’. It’s a very interesting book. It makes the case that, for example, because we no longer have wolves in the UK, then you need to kill deer because otherwise they take over. There are ethical issues. What do you do? Do you reintroduce the wolves? Aside from the fact that the farmers don't want them, again humans dictating what goes. There was a balance that had been disrupted there, and therefore now you have the problem, how do you manage herbivores that proliferate. Rewilding, I guess it depends. If you rewild but you don't reintroduce the species that keep the balance, then you are rewilding in a way that is not balanced where one species will take over maybe. These are really difficult ethical issues, and different people give different answers to that. I think if rewilding, maybe then we would need to introduce wolves and bears to the UK, and that would reestablish the wild balance that existed before humans wiped out these species. So, yeah it depends, that's the answer I suppose.

Josie: Given we've got a lecture theatre full of dog fans, I think there's probably quite a few of us who are quite enthusiastic about the idea of a load of wolves running around Milton Keynes. We've got another question over there.

Kevin: Clara thanks very much for an excellent talk. Kevin Shakesheff, PVC for Research and Innovation. I'm really interested if you could say something about the potential for primate computer interaction, and particularly in terms of refining animal experimentation in degenerative diseases. Is there opportunity for this interaction to help the quality of information we can get in medicine?

Clara: Absolutely, I think so. So the whole point is that if you want to study animal cognition for example, and their progression in one way or another and you want to do it through technology, you have to design something that is centred, that is specifically developed for that particular species, because otherwise what you are measuring is not their cognitive performance, but you are measuring the interference that the technology produces on the interaction with the thing. So definitely, Every time you design a piece of technology that is intended for a human or another animal, you need to start really from scratch. You need to start looking at what are these animals’ characteristics? What are they, whether it’s sensory, physical, cognitive, what are they? What is the environment in which they operate? What are their social characteristics, social interaction, all of these. You take all of these on board, and then you start, as we were saying, develop, design, prototype, and it is an iterative cycle. So definitely, I think if this was taken on board by people who designed these systems for research, and also I think doing it as ethically as possible in that context is very important. You involve the animals as a participant, not by imposing or forcing the animal, because that doesn't allow you to observe what actually is going on with the animal. But if you do it with the animal’s participation and with respect for the animal and taking into account all the characteristics of the design context of which the animal is part, I think you can do a lot better than sometimes is the case.

Josie: Thank you. Have we got any more online? We are a little bit over time, so I should probably wrap up at this point. Clara, thank you very much.

[clapping]

Thank you everyone. An amazing lecture. I really enjoyed that. We strive for continuous improvement and we welcome your feedback. Please do fill in the feedback form which will come out after the event.

Our next inaugural lecture in the series will be by Naomi Moller, who is Professor in the School of Psychology and Counselling in the Open University's Faculty of Arts and Social Sciences. Naomi is going to talk to us about reflections on being a psychotherapy researcher, and that will be on 26^th September at noon. So hoping to see some of you there. Details will be on the OU Research website soon, and you can register there. So thank you everyone for joining us today and for always supporting the OU. For those of us here in person in the auditorium it is time to celebrate so please join us downstairs. Have a good evening everyone. Thank you very much. Clara.