After spending a lot of time studying how people design, use and think about technologies for animals, Dirk van der Linden from Northumbria University started to ponder how to improve data-driven technologies that inform us about animals. While Dirk kept hearing the phrase “data, data, data”, he began to wonder what people need the data for and what insight the information has into what the animals are doing and feeling. From this, Dirk began to believe that the focus of technology shouldn’t be for animals to interact with but instead to change the interactions between animals and humans. To model this, he used an engineering approach creating the model Interspecies Information Systems (IIS). To dig into this further, we interviewed Dirk about his work.
Could you briefly explain your model and how people can apply this to their work?
So the conceptual model is pretty simple, showing just how human and animal actors use technology that in turn provides data-driven suggestions for “interspecies interventions,” which is just a fancy way of saying the computer told you to do this or that to an animal. If you’ve got a FitBark, it might be that it’s telling you to get off your ass and walk your dog. If you’ve got a farm, it might be telling you your cows are spending too much time lying around and to change their routine. So the most significant thing the conceptual model does, and how people can apply it, is that it helps us expand the scope of whatever kind of technology for animals we are thinking about. If you’re designing a new piece of tech, don’t just think about how the animal interacts with it, what they could do, but raise the scope to see how it works entirely—how does the data that this tech captures inform other (human?) stakeholders, if at all? How does it enable actions towards either species? How does it fit into external processes?
Does your model apply to any particular animal?
No, these kinds of information systems arise for wildly different types of animals—from companion animals like dogs and cats indeed, but just as well for apes in zoos, hordes of mongoose and other wildlife, dairy cattle in farming, and so on. Of course, you will see many interspecies information systems currently for dogs, simply because there is a decent market for dogs where people are willing to buy these kinds of technologies. But the sheer size and demand of such technology in farming predates the companion animal market quite a lot, so it’s a matter of business and economy as well.
Why do systems that capture data from an animal need to share data with humans?
Because it’s human-made technology, it’s technology to make us, as human beings capable of being ‘better’ to animals, whatever that means -whether realising the impact of your walks on your dog’s fitness and health, understanding the impact we as a society have on wildlife populations, or, realistically speaking, how to be better at the moment to avoid stress to make their meat tastier, or their milk more productive. Hopefully, the outcomes are good for animals, but humans need to act on them to make those outcomes a reality.
You mention the benefits of IIS is to enable information flow across species, but how does this benefit the animals using the systems?
Like above, the benefit to the animal actors comes through the informed interspecies interventions. Your dog, of course, won’t know that the activity tracker on their collar is capturing data that is processed and sent to you to quantify their physical activity and let you tailor their exercise, diet, and so on. But they will get the effects from the interventions that these information systems suggest where otherwise people might act in the dark or based on generalised advice. So in that sense, this increase in personalised understanding and tailored suggestions of how to best act to animals will benefit them much more than us not having that data.
Why is an engineering approach important to animals using computer systems?
Because it’s all well and nice to design and prototype the fanciest technology we can think of, but in the end, it’s the technologies that we can build that have a shot at making it to the market. That doesn’t just mean knowing how to make the hardware and software’ work’ in our labs, but to engineer the entire ecosystem around it, including viable business models. If we don’t do that, most of our designs will remain in labs and never make it out to the real world, where lots of companies are already getting tech for animals into the hands of consumers.
So, as animal-computer interaction researchers, I think that if we want to impact making technology that improves human-animal relationships, practical engineering approaches are vital to get technology out there in the real world.