Utilising dog-computer interactions to provide mental stimulation in dogs especially during ageing

Abstract

Aged dogs suffer from reduced mobility and activity levels, which can affect their daily lives. It is quite typical for owners of older dogs to reduce all activities such as walking, playing and training, since their dog may appear to no longer need them. Previous studies have shown that ageing can be slowed by mental and physical stimulation, and thus stopping these activities might actually lead to faster ageing in dogs, which can result in a reduction in the quality of life of the animal, and may even decrease the strength of the dog-owner bond. In this paper, we describe in detail a touchscreen apparatus, software and training method that we have used to facilitate dog computer interaction (DCI). We propose that DCI has the potential to improve the welfare of older dogs in particular through cognitive enrichment. We provide hypotheses for future studies to examine the possible effects of touchscreen use on physiological, behavioural and cognitive measures of dogs' positive affect and well-being, and any impact on the dog-owner bond. In the future, collaborations between researchers in animal-computer interaction, dog trainers, and cognitive scientists are essential to develop the hardware and software necessary to realise the full potential of this training and enrichment tool.

Full text

Utilising dog-computer interactions to provide mental
stimulation in dogs especially during ageing
Lisa J Wallis
1,2
Friederike
Range 1,3
Enikő
Kubinyi 2
Durga
Chapagain1
Jessica Serra
4
Ludwig Huber 1
lisa.wallis@live.co.uk
1 Clever Dog Lab, Messerli Research Institution, University of Veterinary Medicine Vienna, Medical
University of Vienna, University of Vienna, Veterinärplatz 1, 1210 Vienna, Austria
2 Senior Family Dog Project, Department of Ethology, Eötvös Loránd University, Budapest, Hungary
3 Wolf Science Center, Messerli Research Institute, Vienna, Austria
4 Royal Canin Research Centre, Aimargues, France
ABSTRACT
Aged dogs suffer from reduced mobility and activity levels,
which can affect their daily lives. It is quite typical for
owners of older dogs to reduce all activities such as walking,
playing and training, since their dog may appear to no longer
need them. Previous studies have shown that ageing can be
slowed by mental and physical stimulation, and thus
stopping these activities might actually lead to faster ageing
in dogs, which can result in a reduction in the quality of life
of the animal, and may even decrease the strength of the dog-
owner bond. In this paper, we describe in detail a touchscreen
apparatus, software and training method that we have used
to facilitate dog computer interaction (DCI). We propose that
DCI has the potential to improve the welfare of older dogs in
particular through cognitive enrichment. We provide
hypotheses for future studies to examine the possible effects
of touchscreen use on physiological, behavioural and
cognitive measures of dogs’ positive affect and well-being,
and any impact on the dog-owner bond. In the future,
collaborations between researchers in animal-computer
interaction, dog trainers, and cognitive scientists are essential
to develop the hardware and software necessary to realise the
full potential of this training and enrichment tool.
Author Keywords
Dog computer interaction; Touchscreen; Animal welfare;
Motivation; Learning; Senior; Dogs.
ACM Classification Keywords
H.5.m. Information interfaces and presentation (ACI):
Miscellaneous
INTRODUCTION
Improving the welfare of captive, domestic or wild animals
is recognised as an important aim of Animal-Computer
Interaction (ACI) [25]. Since animal welfare scientists
realised that welfare problems in animals can be better
addressed with a greater understanding of how animals feel,
there has been a surge of interest in studying animal
sentience [7]. It is universally accepted that animals feel pain
and can suffer; however, well-being is not just the absence
of pain and fear, but is predominantly the presence of
positive affects. Studies in humans have determined that
happiness is promoted by both positive emotions and
positive activities [49]. Positive affect is difficult to measure
in animals, however, evidence from recent studies show that
animals living in enriched environments can benefit from the
creation of situations where there is anticipation of positive
rewards, by promoting play, and opportunities to collect
information, such as in problem solving tasks, which results
in positive physiological and behavioural reactions [12,
32,35]. When these positive emotional experiences are
sustained or repeated, a global state of “well-being” may
ensue, which could help to improve health, and give the
animal a better quality of life [7].
In the UK, nearly one quarter of all households have a dog
[55]. Britain’s spent a record-breaking £7.16bn on their pets
last year, a growth of 25% since 2010 [40]. For the majority
of dog owners, their dogs’ health and well-being is important
as they are considered family members. One particular
section of the market, which is often ignored, is the ageing
dog. As pet dog numbers increase in the UK, so does the
number of old dogs living in our households. The age at
which a dog enters the senior phase of their life varies
according to breed and size, but most dogs can be considered
senior from between five and 10 years of age. As part of the
normal ageing process, senior dogs suffer from a reduced
metabolism and an increase in the occurrence of arthritis and
joint issues, resulting in reduced mobility and activity levels,
which can affect their daily lives, and also influence the dog-
owner bond [3,22].
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.
Copyrights for components of this work owned by others than the author(s)
must be honored. Abstracting with credit is permitted. To copy otherwise,
or republish, to post on servers or to redistribute to lists, requires prior
specific permission and/or a fee. Request permissions from
Permissions@acm.org. © [Lisa Wallis] [2017]. This is the author's version
of the work. It is posted here for your personal use. Not for redistribution.
The definitive version was published in ACI2017, November 21–23, 2017,
Milton Keynes, United Kingdom. Publication rights licensed to ACM.
ACM ISBN 978-1-4503-5364-9/17/11
https://doi.org/10.1145/3152130.3152146

Previous research has determined that dog personality
changes over the course of a dog’s lifetime [20,45,51], which
can lead to changes in dog and owner demographics, such as
in the case of when a dog begins to show signs of ageing.
Reductions reported by owners in their dogs’ personality
traits of trainability, and activity/excitability that occur with
increasing age result in a decline in owner positive attitude
towards their dog, and in turn, a reduction in the amount of
time the owner spends together with their dog in activities
such as walking, playing and training [59]. It is quite typical
of owners of older dogs to reduce all activities with their dog,
since their dog may no longer seem to need or want that type
of stimulation, as there is often a large increase in the time
the dog spends sleeping, and/or inactive during the day [15].
Additionally, the owner’s attitude to ageing in dogs may also
influence how much time they spend active with their dog,
such that if they believe that a dogs golden years should be
spent in quiet and relaxation, then they are likely to reduce
activities with their dog regardless to its ability to take part
in those activities.
Numerous studies have documented the benefits of physical
activity and cognitive enrichment on the performance of
laboratory dogs in different cognitive tasks, and the effect is
particularly strong in aged dogs [34]. There is also evidence
that lifelong training experiences in pet dogs (measured via
owner questionnaires) have the potential to maintain
cognitive function in aged dogs, in a similar way to higher
education in humans. Such that dogs with a high level of
lifelong training perform better in problem solving tasks than
novice dogs [27–29,44] regardless of age, and additionally
have higher levels of attentiveness [9]. Dogs can similarly
benefit from repeated exposure to cognitive enrichment. For
example, old dogs with prior experience on discrimination
learning tasks were quicker to learn new discriminations than
dogs with no such experience [33]. All of these studies point
to the fact that ageing seems to be slowed by mental and
physical stimulation, and thus stopping these activities might
actually lead to faster ageing in dogs.
To address the possibility of a reduction in mental and
physical stimulation in (aged) pet dogs caused by changes in
lifestyle, personality, and mobility status, in this paper, we
explore the potential of touchscreen technology to improve
dogs’ positive affect and novel motivational experiences
through cognitive training. Below we discuss how boredom,
learning opportunities, and motivational changes in dogs can
influence their positive affect and well-being, as well as
detailing the possible benefits of various types of
enrichment.
BACKGROUND
Captive and domestic animals are often passive recipients of
stimulation, rather than having choice and control over their
experiences and behavioural options [61]. For example, for
some social animals, being confined alone for long periods
leads to boredom. Signs of boredom include increased
drowsiness with bouts of restlessness, avoidance and
sensation-seeking behaviour [8]. Captive animals lacking
sensory or cognitive stimulation (such as when exercise,
exploration and/or learning opportunities are reduced) have
weakened neural pathways, which can result in their brains
becoming physically smaller [63]. Some of these boredom
behaviours have been described in dogs that lack physical
and mental stimulation [31]. By providing dogs with
cognitive and environmental stimulation, their quality of life
can be improved, and the prevalence of abnormal behaviours
may be reduced.
Previous research has shown that humans and non-human
animals prefer to work for reward, rather than receiving a
reward for “free” [14,30]. There is no doubt that dogs find
food to be rewarding, as reflected in pleasurable responses to
receiving high value food items. However, one study
determined that providing food directly to dogs, without the
necessity to perform a specific action to get it, may reduce
the hedonic value of the food item [56]. McGowan et al. [30]
showed that dogs displayed higher positive affect (as
revealed through eagerness to enter the test room, increased
activity and tail wagging) when they could control access to
a reward through executing an operate task, than when they
could not control access and only expected a reward. They
concluded that opportunities to solve problems, make
decisions, and exercise cognitive skills are important to an
animal’s emotional experiences and ultimately, their welfare.
Problem-solving opportunities have been found to be
intrinsically motivating as shown by evidence that dopamine
is released during learning and memory consolidation [6].
Researchers have suggested that dopamine acts to stamp in
response–reward and stimulus–reward associations that are
vital for the control of behaviour motivated by past
experience [62].
As part of normal ageing in human and non-human animals,
there is a loss of dopamine neurons, which contributes to a
decline in episodic memory [10]. According to the 'NOvelty-
related Motivation of Anticipation and exploration by
Dopamine' (NOMAD) model proposed by Duzel et al. [11],
dopaminergic dysfunction in old age should also be
associated with diminished motivational drive and energy to
engage in exploratory behaviour, as well as mild motor
dysfunction. The model suggests that cognitive training
combined with reinforcement learning principles and
mobility interventions should result in improvements in
cognitive function and memory, as well as motivation,
novelty seeking, goal orientation, stimulus salience and
exploratory behaviour. Recently, evidence to support this
model came from a study that found that cognitive training
using a memory game on an iPad improved episodic
memory, visuospatial abilities, increased engagement and
heightened motivation in people with age related mild
cognitive impairment [48]. In parallel to the rise in interest
in human brain training, recently there has been a surge of
interest in cognitive training and enrichment for dogs, which
can be implemented by dog owners in the home
environment.

Figure 1. Schematic diagram of the touchscreen apparatus, including: Feeder box (containing food dispenser and
computer/laptop), movable doors to block out distractions, and adjustable computer touchscreen. Treats are dispensed through a
tube from the feeder box, or a feeding device such as the Treat & Train can be used to dispense treats at a distance. Top left:
Photograph of the food dispenser used in the studies. Bottom right: Treat & Train automatic food dispenser with remote control.
Dogs’ need for cognitive enrichment can differ according to
their lifestyle and status, for example, working dogs may not
require additional stimulation in their daily lives. However,
for non-working and senior dogs, technology can provide an
alternative or additional method of cognitive training. Many
new dog intelligence toys and training tools are now
available to buy, and more “cognitive” training methods
using positive reinforcement are being promoted.
Additionally, many Apps have been developed for use by
dog owners to help improve dog welfare, and one has even
been produced for use with dogs, “Game for Dogs”. An
owner, who provides their dog with the use of intelligence
and manipulative toys, might increase the dog’s positive
affect, due to the fact that the dog can “work” for the reward,
which increases the perceived value of the reward. However,
these types of toys quickly become less interesting to the dog
once they learn how to successfully operate them.
One other way of increasing dogs’ positive affect and well-
being is owner-dog play. There is evidence that many dogs
find playing with the owner rewarding, but not all dogs like
to play, and play levels are known to decrease in senior and
geriatric dogs [46,47]. The quality and type of play is mostly
determined by the owner, and can include authoritarian
commands, different levels of enthusiasm, praising, and
petting, as well as physical manipulation, which the dog may
find aversive or pleasant depending on the individual.
Differences in the behaviour, mood and motivation of the
owners can influence the dogs own motivation and behaviour
during play [17]. Not all owners know how to engage in play
with their dog, or are even motivated to play with their dog
at all. Thus, although play can be beneficial in reducing stress
and improving the dog-human bond in some dogs, it is not
suitable for all dogs, especially older dogs, and those that
have mobility issues.
One type of technology that can be implemented for use with
pet dogs, and that has already been used in humans is
cognitive training utilizing games played on touchscreens
and iPads. The power of the touchscreen as a training tool is
in its flexibility, reliability and controllability, and in its
ability to provide novel motivational experiences. The
number of cognitive training possibilities are limitless, as the
stimuli (clipart, photo and even video), acoustic feedback,
reward type, and cognitive paradigm tested can all be varied
[53]. Unlike humans, which can be surprisingly inconsistent
in their behaviour (e.g. when deciding when to praise or
reward dogs, and when giving commands (tone of voice)),
the touchscreen will always give consistent immediate audio
feedback when the dog makes a correct choice, accompanied
by a food reward. The dog learns that the choices it makes
dictate the feedback they receive (positive or negative),
thereby giving a measure of controllability to the dog.
We hypothesize that the touchscreen procedure helps to
create a state of pleasant anticipation in the dog. From the
work of Gregory Berns, we know that there are striking
similarities between dogs and humans in the functioning of
the caudate nucleus, an area of the brain that is associated
with pleasure and emotion. fMRI studies in awake
unrestrained dogs have revealed positive and consistent
responses in the caudate nucleus to objects and stimuli that
dogs liked [4,5]. This means that when the dog begins to
learn to associate the touchscreen and stimuli with a positive
reward, the stimuli and the apparatus itself can create the
state of anticipation of reward even without the presence of
food.

Within the ACI literature, so far only Zeagler and colleagues
have focused on designing a touchscreen interface for use
with dogs [64,65]. They designed a system for service dogs
to relay emergency information about their handlers from a
home or office environment. Therefore, they did not focus
on the touchscreen as a method of cognitive enrichment;
however, they did test several different methods of training
dogs to interact with screens to produce a preliminary
foundation for touchscreen “best practices”, which we have
discussed more in the methods section (see below).
TOWARDS NOVEL EXPERIENCES FOR OLDER DOGS
For too long the old saying “You can’t teach an old dog new
tricks” has been prevalent in society, despite numerous
studies proving the contrary [23,58,60]. We set out to
quantify dogs’ cognitive abilities utilising the Vienna
Comparative Cognition Technology (VCCT), a touchscreen
interface specifically adapted for the use of pet dogs [53].
Owner and dogs participated voluntarily, and dogs were
trained using positive reinforcement. Since 2006, Ludwig
Huber and his colleagues at the Clever Dog Lab have trained
over 200 dogs to use the touchscreen, and have pioneered
touchscreen research in learning, memory, and logical
reasoning abilities of pet dogs [1,18,37,43], and many other
species (such as pigeons [52], marmosets [21], Kea [39],
ravens, tortoises [36], and pigs). Since then, dog cognition
labs around the world have begun to use this technology as a
way to tap into the cognitive capacities of dogs [66,67].
We propose that the use of touchscreen technology has great
potential to improve the quality of life particularly of aged
dogs, by providing an opportunity to participate in a
cognitive enrichment program, which can be tailored and
adapted for the use of senior and geriatric dogs. By providing
an activity that dogs can participate in, which relies almost
exclusively on repeated positive reinforcements, and
problem solving opportunities, we speculate that the
continued use of the touchscreen could result in an increase
in dogs positive affect and motivation. We used the dogs’
behaviour as reported by the owner and trainer, as a non-
invasive indicator of welfare and positive affect, and to
ensure that the user requirements and experience (UX) were
evaluated. Although we did not measure physiological
changes during touchscreen training and testing, an increase
in affect can be reflected in dogs’ motivation to continue
participating, and additionally the owners’ and the trainers
report of the behavioural responses of the dogs during
training, and any changes in the dog’s personality and/or
willingness to participate in activities in the home
environment.
Here we present a methodological contribution, which aims
to detail the necessary hardware and the different training
techniques, which we have so far used to facilitate dog
computer interaction (DCI) in our labs. We emphasize the
changing needs of aged dogs (such as reduced mobility) and
how this may influence the training and testing procedure.
We will also discuss the various methods that can be used to
measure the impact of touchscreen training on dogs’ positive
affect and well-being, and the dog-owner bond in the future
applications section. It is our hope that collectively we can
contribute to designing technology to improve the lives
particularly of aged dogs. We anticipate that this paper will
start a dialogue between different institutions and lay the
foundation for future collaborations.
THE TOUCHSCREEN APPARATUS
The touchscreen apparatus consists of a laptop, a 15” TFT
computer monitor that is mounted behind an infrared
touchframe, and a feeding device that distributes treats
(Figure 1). An infrared touchframe was chosen as the best
option for use with dogs, since it allowed for a level of
moisture, and saliva from the nose presses of dogs, whilst
still functioning. However, dogs with excess saliva may
result in the touchframe becoming unresponsive; therefore,
the screen should be wiped regularly to avoid this occurring.
The monitor and touchframe can be slid up and down to
adjust to the height of the dog. The centre of the screen
should be located at the dog’s eye level (Figure 2).
The feeding device was designed and built by Wolfgang
Berger from the Messerli Research Institute, and contained a
wheel with 32 holes, which rotated to release a single treat
when the dog touched the correct stimulus. Since this system
was complicated to design and make, required regular
maintenance, and was limited in the number of treats that can
be dispensed, several additional options exist regarding
feeding devices, which enable a more multi-functional
approach. The “Treat & Train”, which is relatively
inexpensive, and commercially available from PetSafe,
utilises a remote control that the owner/trainer can press to
dispense a single treat, if the dog makes a correct choice [68].
Another available dispenser is the “Pet tutor”, which has the
additional advantage of Bluetooth connectivity [69]. The
distance between the screen and the dispenser can be varied.
It would also be possible to use the device designed by
Wolfgang Berger as a moveable dispenser if it is not
integrated within the touchscreen Feeder Box. This has the
benefit of causing the dog to move away from the screen to
obtain the reward, and may help to give the extra seconds
required for the dog to view the screen and make a correct
choice. When utilising this feeding method, the size of the
touchscreen apparatus can be substantially reduced. For
example, the screen and touchframe can be mounted onto a
wall, and any cables covered, to ensure the dog could not
gain access to them. If using a laptop to run the software, it
should be placed out of reach of the dog, on a nearby shelf or
table, so as not to confuse the dog with access to two screens.
Movable doors or screens are located at the front of the
touchscreen, which can be folded out to create a “testing
niche”, which helps to prevent distraction from the external
environment, and also serves to position the dog in the ideal
location to utilise the touchscreen (Figure 2). Many dogs
approached the touchscreen from the side in the initial
training, and not head on. This could cause a side bias to
develop; therefore, the dogs need to be guided into the testing

niche to position them centrally and to allow optimal viewing
of the stimuli.
Both the “Treat & Train” and the “Pet tutor” can be adapted
to integrate them with the computer, which would allow the
automatic triggering of the feeding device once a correct
choice has been made by the dog, (by touching the correct
stimulus on the screen). When utilising this method, the
pressing of a remote device by the trainer would no longer
be necessary. Please note that on occasion these devices may
fail, which may result in food becoming jammed in the
feeder, and/or no food reward being dispensed. If this is a
regular occurrence, then the trainer can have some extra
treats in a food pouch, which she/he can quickly drop into
the dispenser food bowl when necessary, before fixing the
device. Careful consideration should be given to the type of
food used as a food reward. The dietary requirements of the
individual dogs should be determined, and in most cases the
dogs’ usual dry food, or semi moist food, can be used. In the
case of over-weight dogs, a dry food with reduced calories
can be utilised, or a low calorie training treat. In our
experience, some of the dogs were too quick to eat the dry
food in their haste to return to the touchscreen, which
sometimes resulted in occasional choking. We experimented
with the consistency of the reward, and found that a chewy
or semi-moist option worked better with these types of dogs.
Using wet food as a reward in automatic feeders is not
currently possible; however, for dogs that require more
motivation, small pieces of cubed hard cheese and/or hotdog
can be mixed in with the dry food. Please note that these
softer foods may become jammed in the devices, apart from
the feeder designed by Wolfgang Berger, which was set up
to allow the use of different food types. On training days, the
dog’s food allowance should be lowered to incorporate the
amount of food used during touchscreen training.
Additionally the dog should not be fed for at least two hours
before training, to ensure sufficient motivation.
SOFTWARE AND PICTURE STIMULI
At the Clever Dog Lab, and the Wolf Science Center, to
present the stimuli and record the dogs’ responses, we use a
software program called CognitionLab by Michael Steurer
(version 1.9; see ref. [53] for detailed description). The
software presents picture stimuli (in jpeg or bmp file types)
according to the users specifications. We used a “First-
Contact” touch system, where the first contact with the target
counts, even if it is not the first impact with the surface. This
method was also found to be the easiest to learn for dogs by
Zeagler et al. [65]. The software allows great flexibility in
modifying inter-trial intervals, stimulus positions, auditory
feedback, background colour, use of correction procedures,
and presentation time. Data are logged into a single text file
per subject, and contains such information as, which stimulus
was touched first, the precise location of the touch, and
additionally undefined touches (areas the touchscreen was
touched that did not contain stimuli) including all precise
timings of the touches. In the near future, CognitionLab will
become open source, and therefore available for all to use.
One other open-source program available that could be used
is called OpenSesame. Recently, at the Clever Dog Lab, our
newest touchscreen apparatuses are run utilising open source
software called “DogTouch” designed by Messerli Research
Institute Computer Technicians Michael Pichler and Peter
Füreder. We base our newest hardware on commercial-level
components and custom solutions. We integrate Arduino, an
in-expensive open-source electronics platform based on
easy-to-use hardware and software, into the design.
Additionally, we are experimenting with microcomputers
(e.g. Raspberry Pi), the emerging standard widely accepted
in the do-it-yourself “maker” community. This will lead to
inexpensive hardware becoming accessible to a large
community, capitalizing on the fast-growing "ecosystem" of
companies serving the maker community.
Figure 2. A dog standing in the “testing niche” (moveable
doors folded in) demonstrating the correct position for the dog
to best utilise the touchscreen apparatus.
The stimuli displayed on the touchscreen consist of jpeg clip
art images obtained from the internet presented on a white
background. The stimuli should differ in colour, global
outline, and internal features. This will allow the dog to more
easily discriminate them. When training aged dogs, we found
the optimum size of the stimuli to be 200 by 200 pixels,
which is equivalent to about 2 inches in size. However, if a
dog has mobility problems, and touching precisely on a small
square image is more of a challenge, the stimulus size can be
increased to 300 x 300. Please note these picture sizes were
presented on a 600 x 800 pixel screen, so the resolution of
the screen must be taken into account when sizing stimuli.
For comparison, Zeagler et al. in their touchscreen
methodological study used stimuli at a size of at least 3.5
inches [65].
SUBJECTS
So far, around 265 dogs, and 20 wolves were trained to use
the touchscreen in several different studies in several
different labs, including the Clever Dog Lab and Wolf
Science Center in Austria, and the Family Dog Project in
Budapest. Most of the dogs were pet dogs living with
Austrian or Hungarian families, however, 20 dogs were
raised, socialized, and kept in enclosures at the Wolf Science
Center in a similar way as to the 20 wolves housed there. One
hundred pet Border collies (aged from 5 months to 14 years),
and 115 dogs (aged 6 years and over), from different pure

breeds and mixed breeds were trained at the Clever Dog Lab.
Thirty pet dogs of various breeds and ages were trained at the
Family Dog Project, and 20 dogs and 20 wolves were trained
at the Wolf Science Center, Austria.
TRAINING PROCEDURE
Here we will focus on the preliminary steps necessary to train
a dog to interact with the touchscreen. The dog receives a
training programme consisting of several phases of
progressive complexity. The goal of the auto-shaping and
pre-training procedures is to familiarize dogs with the
touchscreen apparatus and the food dispenser (Phase 1), to
teach them to touch a stimulus on the screen, first at a fixed
location (Phase 2), then at varying locations (Phase 3). Then
finally, to select a picture from two or more to obtain a
reward (discrimination training, Phase 4). Only then can the
dogs start to solve more difficult cognitive tasks, which can
further examine their learning, memory and logical
reasoning skills [1,43,60]. Students and research assistants,
some of which had considerable previous dog training
experience trained the dogs at our labs. All were briefly
instructed in the basics of the different training techniques,
however, needed some practice before perfecting their
training skills. Therefore, when we refer to a “trainer” in the
text, we denote someone who has an understanding of how
to train dogs, and a good knowledge of the individual dogs
being trained.
Phase 1 Familiarization with the touchscreen apparatus
and the food dispenser
Owners brought their dogs to the lab once a week and
participated in three to four sessions, over a half-hour period,
with short breaks in between sessions. Initially the trainer
had to help the dog using a variety of techniques (such as
shaping, target training, and luring), to approach the
apparatus and the screen, which is of course, not a natural
behaviour for the dog, and additionally the dog needed to
learn how to use the feeding device. We found that during
this early stage, the movable doors at the front of the
touchscreen should be positioned in a wide-open position, so
that the dog and the trainer can approach the front of the
apparatus unimpeded. The quickest method to familiarise the
dog with the apparatus is to use luring, which consists of the
use of food to guide the dog into a desired position or
behaviour. Liver sausage is a treat that most dogs enjoy, and
can be obtained as a paste in a handy tube dispenser. Cream
cheese or peanut butter can also be used for fussy eaters, or
dogs with specific food allergies. Initially, the paste should
be smeared on the touchscreen to attract the dog to the
apparatus, this step is especially important in fearful dogs, as
the apparatus itself as a novel object may be potentially
scary. If the dog is familiar with the owner/trainer using
shaping when training, and is used to offering behaviours and
interacting with objects, as well as knows the “touch”
command (used specifically to ask the dog to touch an object
with its nose), then this is often the quickest method of
training the dog to approach the screen. Shaping involves
breaking down a behaviour into tiny increments, and
reinforcing the dog at each incremental step until you've
achieved the full behaviour. Here the dog is rewarded for
approaching the apparatus, then for sniffing the apparatus,
then touching, then touching specifically the screen. The use
of a clicker device if the dog is familiar with it can speed up
training. Generally, we found the shaping technique to work
well for the Border collies, as most were already highly
trained and familiar with the use of the clicker, and shaping
technique. For these dogs, many of them only required one
or two visits before performing reliably, and moving onto the
next training phase. This method was not suitable for many
aged pet dogs, which had no such experience with shaping
training methods.
Once the dog is familiar with the touchscreen apparatus, the
dog should also be accustomed to the feeding device. The
feeding device is necessary to avoid that the dog begins to
focus too much on the trainer during training and testing.
Some dogs that have never worked with a feeding device
may paw or chew the device in an attempt to open it to get at
the food they can smell (and in some models, see), inside. In
most cases, a short training session with the feeding device
is necessary for the dog to learn that food will only be
available when the trainer presses the remote (or is triggered
automatically by the software); the device emits a beep, the
motor turns and the food reward drops into the feeding bowl.
For noise sensitive dogs, the volume of the beep can be
lowered, or even turned off, and can be slowly increased as
the dog becomes accustomed to it. Some dogs may find it
very hard to inhibit standing/lying next to and/or
manipulating the feeding device to the extent that it can be
very hard to lure them away. In these specific cases, we
recommend that the feeder be placed inside the touchscreen
housing, and only the dish at the bottom (where the
dispensed treat appears) should be available to the dog. Once
the dog has no problem with approaching the apparatus, and
is familiar with the feeding device the software program
should be initiated. Some dogs needed only one visit to reach
this point, others needed two.
Phase 2 Touch a stimulus on the screen (fixed location)
In the approach training, which consists of the presentation
of a single stimulus, when the stimulus is touched, the
infrared light grid on the touchframe is interrupted, which
triggers an acoustic signal and, in the case of an automated
system, the delivery of a food treat. The training requires that
the dog learns the association between touching the picture –
and gaining a food reward. From our experience of testing
dogs with many different picture stimuli, and from the dog’s
visual capabilities, we have determined that a stimulus with
a roughly circular global shape and blue and yellow in colour
is particular eye catching for dogs, and serves as a good
starting stimulus. For dogs familiar with shaping and the
touch command, in a final step, the dog can be rewarded for
touching the stimulus on the screen. The finger can be used
to lure the dog to the screen, and to get it to touch the screen
in the correct location. Luring is by far the easiest method of
teaching the dog to touch the stimulus. However, it took us

some time to perfect the technique, and avoid that the dog
becomes too focused on the trainer, or on the hands. Luring
and shaping were also used in Zeagler’s study to train service
dogs to use touchscreens in a tapping task [65]. Zeagler
found that the luring approach was less effective than
shaping, as the dogs did not associate the completion of the
task with the reward, as they focused instead on the screen
where they expected the reward to appear. To help combat
this problem we recommend the following procedure as the
quickest and most efficient method to train aged dogs to
utilise the touchscreen.
Once the single stimulus (here the blue flower) appears on
the screen, the trainer stands to the left side of the
touchscreen (facing away from the screen), takes a blob of
paste onto his/her right index finger, and then with the left
index finger should reach over and touch the edge of the
touchscreen. Next, the trainer uses his/her right index finger
to smear the paste directly on top of the stimulus. By
touching the edge of the screen with the left hand, the trainer
ensures that when the right hand touches the stimulus, it will
not trigger the touchscreen to activate a correct choice. After
placing the paste on the stimulus, the trainer withdraws the
right hand, and then the left, and holds their hands behind
their back, to avoid the dog being distracted by the presence
of the nice smelling fingers. Once the dog begins to lick off
the paste from the screen, the tongue will activate the
touchscreen, and the computer will register a correct choice
and produces a beep, and the stimulus will disappear. At
which point the feeding device is activated (in the case of an
integrated system) or the owner/trainer should activate the
feeding device via the remote control. In the latter case, the
timing is crucial, and the device should be activated as
quickly as possible upon hearing the beep from the computer.
Effectively, in this phase the dog is rewarded twice, once
when licking off the paste and the second time by the feeding
device. Therefore, we can avoid that the dog focuses only on
the screen or the reward, but learns that the action of touching
the screen also results in a reward from a separate location.
Now that the dog is familiar with the feeding device, the best
position for the device can be determined during this phase,
taking into account the physical ability of the dog. For some
dogs, placing the feeder a few meters away can provide some
much-needed additional exercise, helping to increase the
dog’s activity level, and has the added benefit of giving the
extra seconds required for the dog to view the stimulus
before reaching the touchscreen, and therefore may touch
more precisely. Other dogs, which have specific mobility
problems, can be helped by placing the feeder close by or
even under the touchscreen, and the device can additionally
be raised up, so the dog does not need to lower its head to the
ground. The touchscreen can be operated by the dog from a
sitting, or even a lying position if necessary, for those dogs
with chronic pain and mobility issues.
In the first instance, it may be necessary for the trainer or
owner to point to the location of the dispensed treat, to assist
the dog in finding it. While the dog is eating the food reward,
the trainer can immediately repeat the process of touching
the screen with the left hand, and applying the paste with the
right onto the new stimulus presented on the touchscreen.
The dog will begin to learn to associate the beep and the
sound of the feeder with the food treat. However, some time
is necessary for the dog to learn the routine of first licking
the screen, and then looking for the food reward. In general,
most dogs were able to pass this stage in several visits, but
occasionally some older dogs needed three visits.
Phase 3 Touch a stimulus on the screen (varying
locations)
In a slight change to the approach training, the position of the
stimulus is randomly alternated between the left of centre
and right of centre positions. The same training technique as
detailed in Phase 2 can be implemented. Once the dog learns
to touch precisely the stimulus, the trainer can slowly reduce
the amount of paste that is applied to the screen, until the
paste is no longer necessary. At this point, the dog generally
switches to a nose press, instead of a lick. The movable doors
at the front of the touchscreen should be slowly closed, to
ensure that the dog stands correctly, and to minimize outside
disturbance. Several visits may be necessary to reach this
stage, but if the dog becomes confused, and does not offer
the correct behaviour after prompting, the use of the paste
can be reinitiated until the dog reliably executes the touch
action, and immediately goes to the dispenser to receive a
food reward. By the end of the Phase 3 training, the dog
should successfully complete one session (30 trials) with no
help from the trainer. In our experience, the aged dogs
needed an average of three visits to reach this criterion.
Therefore, in total from Phase 1 to completing Phase 3 aged
dogs required around seven visits (range = 1 – 15). All dogs
were able to complete the approach training, and then moved
on to the next training phase, a two-choice discrimination.
Phase 4 Discrimination training
In the final training step, using a forced two choice procedure
(which just means that the dog must press one of two possible
stimuli), the software presents one positive picture stimulus
(S+) and one negative picture stimulus (S−), positioned
randomly on the left and right side from trial to trial (for an
example, please see Figure 2 and 3). When the positive
stimulus is touched, both stimuli disappear, a short tone is
emitted by the computer, and a food reward is provided. If
the wrong stimulus is touched (S−), both stimuli disappear, a
short buzz sounds, and a red “time out” screen is presented
for three seconds. In this case, a correction trial is
immediately initiated: the stimuli of the previous trial are
presented again in the same position as previously. If the dog
makes a correct choice, the trial terminates and results in a
food reward and the presentation of a new trial. A second
incorrect response results in a further correction trial. After
each trial (with the exception of correction trials), an inter-
trial interval of two seconds is initiated where an empty white
background is presented. In order for the dog to learn the
difference between the positive and negative stimuli, both

positive feedback (tone and treat) and negative feedback
(short buzz, red screen and three second time out) are
necessary.
Figure 3. A Border collie working on the touchscreen in the
two choice discrimination
The dogs’ task is to discriminate reliably between the two
stimuli. We set an arbitrary criterion (20 or more correct
choices in 30 trials (66.7%) in four out of five sessions) for
moving onto the next training phase. When this task is first
initiated, the dogs may become confused and unsure of how
to respond. Indeed many dogs tried to touch in the middle
between the stimuli, or tried to touch both by sliding the nose
across the screen. The first few times the dog touches the
negative stimulus can cause a measure of frustration to the
dog when the action does not produce the expected food
reward. Help from the trainer/owner is often necessary to get
the dogs to continue touching the stimuli, for example, verbal
encouragement to approach the screen and touch, and
occasional pointing. The dog may also attempt to use
strategies other than discrimination to solve the problem,
such as always choose the stimuli on the left, and when
incorrect choose the right stimuli, or a win-stay lose-shift
strategy. Some dogs can benefit from extra time to process
the stimuli, so here the feeder may be moved further away
from the screen, or the trainer may use their arm to prevent
the dog from entering the testing niche for a few seconds, to
allow the dog to slow down, and view the stimuli. However,
the trainer should not block the dog with their arm, or hold
the dog by the collar/harness whilst watching the stimulus
presentation. It is our experience that the trainer may
unconsciously release the dog when it is attending to the
correct stimulus, and thereby the dog can use subtle cues
from the trainer to solve the discrimination, without actually
learning the correct contingencies of the stimuli.
Eventually the dog will learn the discrimination after a
certain number of sessions elapse, depending on the dogs
learning abilities. On average, dogs over 6 years took 15
sessions (range = 4 – 40) to reach criterion. Only two older
dogs (out of a total of 130 dogs aged above 6 years) failed
this training stage, so it is well within the capacities of the
majority of senior dogs. Our research indicates that dogs’
ability to discriminate stimuli on the touchscreen improves
with the training of new additional stimulus pairs
(publication pending). However, the dogs are heavily
influenced by the characteristics of the stimuli themselves,
such as brightness, colour, contrast, and luminosity. We are
currently researching, which stimulus properties the dogs
attend to in two-choice discriminations. The touchscreen
apparatus and software offers an almost limitless opportunity
to examine the cognitive capacities of dogs. Once they
master the pre-training, then many of the dogs moved onto
more difficult tasks, such as categorization [43], face
discriminations [41], emotional discriminations [2,37],
numerical discriminations [42], and inference by exclusion
(a kind of fast mapping, exemplified by Ricoh, the Border
Collie [19]) [1,60].
OWNER FEEDBACK
To assess the user requirements and experience (UX), the
trainers of the dogs spoke to the owners regarding their dog’s
progress as well as their expectations, during every weekly
visit. Initially many owners were sceptical regarding whether
their dogs were capable of learning the touchscreen
paradigm, especially the owners of aged dogs. However, all
the owners were interested in the studies and motivated to
participate. Some owners travelled for over an hour by car to
reach the labs, and some even came twice a week. After
several visits, the owners were often amazed to see how well
their dogs were progressing, and the enthusiasm of their
dogs, when they began to anticipate their weekly training
sessions. Many owners referred to their dogs as computer
geeks and were quick to express the fact that they believed
their dogs enjoyed participating in the study. So much so,
that there was a low drop-out rate (around six dogs in total),
despite the fact that for some dogs, the full training and then
testing in more complex tasks lasted over a year, and owners
were not compensated for participating. The positive
association to the touchscreen is so strong that on several
occasions when the dog was alone (the trainer had stepped
out to answer the phone), and the feeder failed, dogs
continued to work on the touchscreen with no reward until
the end of the session. Additionally, several dogs were
trained on the touchscreen when they were younger, and then
had substantial gaps of between 3 and 7 years, before they
returned to the lab for a new study. These dogs not only
remembered the touchscreen procedure, but also in several
cases recalled the correct stimuli on a discrimination learnt
over three years previously.
Watching the dogs learning on the touchscreen was
illuminating for the owners, as it revealed much about the
dogs’ character regarding the strategies they used, and had
the added bonus of tiring the dogs out mentally. After the
training when the dogs returned home, many of them fell into
a restful sleep, similar to that after a bout of exercise. For
many owners this mental tiredness was a new concept, and
stimulated them to try other mind games to play with their
dogs on days when the dogs were not trained on the
touchscreen. Owners received a certificate when their dog
completed the training, and many of the owners framed it and
placed it on display in their homes. We have no doubt that
participating in the studies improved the dog-owner

relationship, as reported via personal communication with
the owners. However, we should point out that the dog-
owner relationship was likely already quite positive, as the
study was voluntary, and was likely to attract highly
motivated owners. Studies are currently underway to analyse
owner questionnaires designed to identify dogs’ behavioural
responses to touchscreen training.
THE ETHICS OF DOG-COMPUTER INTERACTIONS
A welfare centred ethics framework in ACI research has
recently been proposed by Mancini [26]. A review of the
aims of ACI (to improve welfare, benefit animals, and
improve the human-animal relationship) as well as the
potential harms and benefits was carried out by Grillaert and
Camenzind [13]. They suggested that there is a need for
greater focus on data collection, the long-term implications
of ACI use, assessments of how it may influence animal time
budgets, and the importance of preference tests. To address
these issues, in the future we would like to implement video
behavioural analysis, owner questionnaires, and dog activity
monitors in long-term touchscreen studies.
We are aware of the fact that for certain personality and/or
breed types, interacting with the touchscreen might induce
anxiety, over-arousal, or other behavioural changes that
might cause welfare issues for the individual or harm the
human-animal relationship [57]. In our experience adverse
reactions to the touchscreen paradigm were very rare,
however, three dogs (1.5% of the sample) showed increased
vocalizations, and/or a measure of frustration (such as
pawing at the touchframe, excessive panting, turning or
walking away from the apparatus) when presented with the
second pre-training step – the two-choice discrimination. We
interrupted training immediately if we detected signs of
distress or undue frustration. After a break, the trainer
attempted to give extra help to these dogs to learn the new
procedure, and reduce their negative behavioural reactions to
the fact that touching one of the stimuli resulted in no reward
(for example, by initially covering the negative stimulus with
a piece of card). If stress signs re-emerged, then continuation
in the program on a later occasion was discussed with the
owner. Therefore, we should emphasize that although the
majority of the dogs showed only minimal signs of
frustration (such as occasional tongue flicks, yawning, or
shaking off), for a small proportion of the dogs, the
touchscreen paradigm is not suitable. Thus, for successful
DCI it is a requirement that the trainer/owner have a good
understanding and recognition of the dog’s needs, stress
behaviours, and learning abilities.
FUTURE APPLICATIONS
So far, the touchscreen apparatus has been used only within
lab environments. Apart from one dog trainer, who holds
workshops on dog-iPad interaction in the United States [24].
However, with the advance of technology, the development
of new applications, and the relatively cheap production of
touchscreen and feeding devices, the touchscreen apparatus
has the potential to be further developed to produce a
working system. An existing laptop could be combined with
a monitor and touchscreen overlay (for example from
Soladapt “Touch Genie” infrared overlay [50]), and a Treat
& Train. The system (not including the laptop) would cost in
the region of £350. The touchscreen apparatus could be
marketed for use within the dog training community.
Trainers at dog’s schools could club together to buy the
equipment necessary to set up their own touchscreen system.
They could offer their services to provide owners with the
opportunity for their dogs to participate in cognitive
enrichment programs, and could run workshops on how to
train dogs to utilise the touchscreen. Owners that are
prepared to spend money on this training tool would also
have the opportunity to purchase the system to improve their
dog’s well-being in the home environment. A rent-to buy
scheme could be implemented, which would increase the
affordability of the hardware, and allow owners to “try out”
the system at home after completing an online training
workshop/seminar on touchscreen training.
Software and application developers should team up with
trainers and cognitive biologists to create new software,
which could allow owners to become citizen scientists. This
means that data from the dogs’ progress on the touchscreen
could be uploaded to a cloud on the internet, which would
allow the data to become available to cognitive scientists,
who could use it to write publications, which would further
increase our knowledge on cognition in dogs. Such a system
is already in place with the popular science-based games
subscription service Dognition [54], that utilises owners to
gather information about their dogs performance in
standardized behavioural tests. One application
programming interface (API) that currently is used to gather
public data for scientific purposes is Fitbark. Fitbark has a
health baseline database of over 200 breeds of dogs from
over 100 countries, which could be used by third party
developers to create new web and mobile apps that could be
geared towards gathering additional data about the positive
effects of cognitive enrichment technology.
The healthy ageing of dogs and the specific needs of the
senior dog is slowly becoming more public knowledge, due
to an increase in the information available, for example
through ageing dog clinics at veterinary surgeries and
physiotherapy centres, web resources, magazine articles and
dog trainers. The touchscreen apparatus could become an
important addition resource to provide cognitive enrichment.
However, additional research is necessary to determine
whether continued use of the touchscreen results in:
 Increases in aged dogs’ positive affect during training (as
measured via owner/trainer questionnaire, video analysis
of behavioural responses during training (tail wagging
and willingness to work), or increases in stress signs
(vocalisations, destructive behaviours and avoidance).
 Increases in motivation, and learning, memory and
visuospatial ability in subsequent touchscreen tasks, and
other cognitive and behavioural tests.

 Changes in dogs’ personality as measured via owner
questionnaire: increases in trainability, activity and
excitability, or behavioural test batteries: increases in
motivation, novelty seeking, exploration, and activity, or
in the daily environment: Fitbark measures of rest/active
and playtime, sleep score, and overall health index.
 Decrease in short-term cortisol measurements and
increases in dopamine during touchscreen training
(indicating low stress and high motivation) combined
with behavioural observations and owner/trainer
questionnaires.
 Positive changes in the dog-owner relationship as shown
by owner questionnaires, and the amount of time the
owner spends active with their dog, which could also be
measured using Fitbark/Fitbit.
In DCI studies, we are heavily reliant on owner
questionnaires, and since dog owners vary in their
experience in understanding and describing dog behaviour,
tool such as the Dog Information Sheet (DISH) [16] can be
utilised, to generate a more informed interpretation of the
dog’s feedback by the proxy-observer when using the
touchscreen technology.
Intervention studies could examine the influence of the
various types of cognitive enrichment (including touchscreen
training, and intelligence and manipulative toys), as well as
physical enrichment (physiotherapy and owner-dog play),
and dietary antioxidant supplementation on successfully
ageing dogs, and dogs with separation related anxiety, and
canine cognitive dysfunction. Finally, dogs that have been
trained to remain motionless in fMRI machines could be
taught the touchscreen procedure to examine how their brain
processes the touchscreen stimuli to provide more evidence
of the cognitive enriching effects of the touchscreen, and to
determine at which point the stimuli properties are encoded
into long-term memory.
There has been considerable interest in producing
technology that can be used by dogs in the home
environment, to help relieve boredom and separation
anxiety, while their owners are away at work [38]. The
touchscreen as it has been described is suitable for use in the
home, however, dogs should be supervised during use, as
problems with the feeder malfunctioning, excessive saliva on
the touchframe, or scratching the touchframe with the paw,
can cause the hardware to become unresponsive and could
result in unnecessary frustration and stress to the dog.
However, once these issues have been addressed, an
improved apparatus should allow unsupervised home use in
the future after individual piloting by the owners.
CONCLUSION
In conclusion, we believe that the touchscreen apparatus and
developed software could potentially improve the welfare of
pet dogs and aged dogs in particular, through cognitive
enrichment. However, further studies are necessary to
determine the effects of long-term touchscreen use on dog
personality, activity levels, and measures of well-being, as
well as any influence on the dog-human bond.
Collaborations between researchers in ACI, dog trainers, and
cognitive scientists are essential to develop the hardware and
the software necessary to realise the full potential of this
training and enrichment tool.
ACKNOWLEDGMENTS
We would like to thank all the owners who volunteered to
participate in the studies, and the students/research assistants
from the Clever Dog Lab: Angela Gaigg, Teresa Marmota,
Manuel Kemethofer, Julia Schößwender, Ina Pohl, Carmen
Mittendrin; Wolf Science Center: Bea Belenyi, Marianne
Heberlein, Christina Mayer, Marleen Hentrup, Rita Takacs;
Family Dog Project: Rita Báji and Barbara Csibra, for
providing assistance with recruiting/training the
dogs/wolves. We are also grateful to Clever Dog Lab
Manager Karin Bayer and team assistants Jennifer Bentlage
and Aleksandar Orlic, and Messerli Computer Technicians
Michael Pichler and Peter Füreder, as well as external
software advisor Michael Steurer. Technical support was
provided by Bence Ferdinandy at the Family Dog Project,
and project management by Márta Gácsi. Furthermore, we
would like to thank our sponsors Royal Canin for providing
funding for this project. LW was additionally supported by
the DK CogCom Program (Austrian Science Fund Doctoral
Programs W1234). Writing was supported by a FWF grant
(project number: P24840-B16) to FR, WWTF project CS11-
025 and the FWF grant P21418 to LH, grant P21418 to LH
and FR. Finally, this project has received funding from the
European Research Council (ERC) under the European
Union’s Horizon 2020 research and innovation programme
(Grant Agreement No. 680040), and by the János Bolyai
Research Scholarship of the Hungarian Academy of
Sciences (to EK).
REFERENCES
[1] Aust, U, Range, F, Steurer, M, and Huber, L. 2008.
Inferential reasoning by exclusion in pigeons, dogs, and
humans. Anim. Cogn. 11, 4, 587–97.
[2] Barber, A, Randi, D, Müller, C, and Huber, L. 2016.
The processing of human emotional faces by pet and
lab dogs: evidence for lateralization and experience
effects. PLoS One 11, 4, e0152393.
[3] Bellows, J, Colitz, C, Daristotle, L, Ingram, DK, … and
Zhang, J. 2015. Common physical and functional
changes associated with aging in dogs. J. Am. Vet. Med.
Assoc. 246, 1, 67–75.
[4] Berns, G, Brooks, A, and Spivak, M. 2012. Functional
MRI in awake unrestrained dogs. PLoS One 7, 5,
e38027.
[5] Berns, G, Brooks, A, and Spivak, M. 2015. Scent of the
familiar: An fMRI study of canine brain responses to
familiar and unfamiliar human and dog odors. Behav.
Processes 110, 37–46.
[6] Bethus, I, Tse, D, and Morris, R. 2010. Dopamine and
memory: modulation of the persistence of memory for

novel hippocampal NMDA receptor-dependent paired
associates. J. Neurosci. 30, 5, 1610–1618.
[7] Boissy, A, Manteuffel, G, Jensen, M, Moe, R, … and
Aubert, A. 2007. Assessment of positive emotions in
animals to improve their welfare. Physiol. Behav. 92, 3,
375–397.
[8] Burn, C. 2017. Bestial boredom: a biological
perspective on animal boredom and suggestions for its
scientific investigation. Anim. Behav. 130, 141–151.
[9] Chapagain, D, Virányi, Z, Wallis, L, Huber, L,… and
Range, F. 2017. Aging of attentiveness in border collies
and other pet dog breeds: The protective benefits of
lifelong training. Front. Aging Neurosci. 9
[10] Chowdhury, R, Guitart-Masip, M, Bunzeck, N, Dolan,
R, and Duzel, E. 2012. Dopamine modulates episodic
memory persistence in old age. J. Neurosci. 32, 41,
14193–14204.
[11] Düzel, E, Bunzeck, N, Guitart-Masip, M, and Düzel, S.
2010. NOvelty-related Motivation of Anticipation and
exploration by Dopamine (NOMAD): Implications for
healthy aging. Neurosci. Biobehav. Rev. 34, 660–669.
[12] Ernst, K, Puppe, B, Schon, P, and Manteuffel, G. 2005.
A complex automatic feeding system for pigs aimed to
induce successful behavioural coping by cognitive
adaptation. Appl. Anim. Behav. Sci. 91, 3–4 , 205–218.
[13] Grillaert, K and Camenzind, S. 2016. Unleashed
enthusiasm. In Proc.Third Int. Conf. Animal-Computer
Interaction ACI, 1–5.
[14] Hagen, K and Broom, D. 2004. Emotional reactions to
learning in cattle. Appl. Anim. Behav. Sci. 85, 203–213.
[15] Harper, EJ. 1998. Changing Perspectives on Aging and
Energy Requirements: Aging and Energy Intakes in
Humans, Dogs and Cats. Am. Soc. Nutr. Sci. 2623–
2626.
[16] Hirskyj-Douglas, I and Read, J. 2016. Using
behavioural information to help owners gather
requirements from their dogs’ responses to media
technology. Proc. Human Computer Interaction, 1–13.
[17] Horváth, Z, Dóka, A, and Miklósi, Á. 2008. Affiliative
and disciplinary behavior of human handlers during
play with their dog affects cortisol concentrations in
opposite directions. Horm. Behav. 54, 1, 107–114.
[18] Huber, L, Apfalter, W, Steurer, M, and Prossinger, H.
2005. A new learning paradigm elicits fast visual
discrimination in pigeons. J. Exp. Psychol. Anim.
Behav. Process. 31, 2, 237–46.
[19] Kaminski, J, Call, J, and Fischer, J. 2004. Word
Learning in a Domestic Dog: Evidence for “Fast
Mapping.” Science. 304, 5677, 1682–1683.
[20] Kubinyi, E, Turcsán, B, and Miklósi, Á. 2009. Dog and
owner demographic characteristics and dog personality
trait associations. Behav. Processes 81, 3, 392–401.
[21] Kurz, D. 2011. Common Marmosets can discriminate
between positive and negative stimuli, but can they also
reason by exclusion? “. University of Vienna.
[22] Larsen, JA and Farcas, A. 2014. Nutrition of Aging
Dogs. Vet. Clin. North Am. Small Anim. Pract. 44, 4 ,
741–759.
[23] Lillard, AS and Erisir, A. 2011. Old dogs learning new
tricks: Neuroplasticity beyond the juvenile period. Dev.
Rev. 31, 4, 207–239.
[24] Macpherson, R. 2013. Wet noses to the touchscreen,
iPads go to the dogs. https://phys.org/news/2013-08-
noses-touchscreen-ipads-dogs.html
[25] Mancini, C. 2011. Animal-computer interaction (ACI):
a manifesto. ACM Interact. 18, 4 , 69–73.
[26] Mancini, C. 2017. Towards an animal-centred ethics
for Animal–Computer Interaction. Int. J. Hum.
Comput. Stud. 98, 221–233.
[27] Marshall-Pescini, S, Frazzi, C, and Valsecchi, P. 2016.
The effect of training and breed group on problem-
solving behaviours in dogs. Anim. Cogn. 19, 571-579.
[28] Marshall-Pescini, S, Passalacqua, C, Barnard, S,
Valsecchi, P, and Prato-Previde, E. 2009. Agility and
search and rescue training differently affects pet dogs’
behaviour in socio-cognitive tasks. Behav. Processes
81, 3, 416–422.
[29] Marshall-Pescini, S, Valsecchi, P, Petak, I, Accorsi, P,
and Previde, E. 2008. Does training make you smarter?
The effects of training on dogs’ performance in a
problem solving task. Behav. Processes 78, 449–454.
[30] McGowan, R, Rehn, T, Norling, Y, and Keeling, L.
2014. Positive affect and learning: exploring the
“Eureka Effect” in dogs. Anim. Cogn. 17, 3, 577–587.
[31] Mcmillan, FD. 2002. Development of a mental
wellness program for animals. J. Am. Vet. Med. Assoc.
7, 965–972.
[32] Meehan, CL and Mench, JA. 2007. The challenge of
challenge: Can problem solving opportunities enhance
animal welfare? Appl. Anim. Behav. Sci. 102, 246–261.
[33] Milgram, NW. 2003. Cognitive experience and its
effect on age-dependent cognitive decline in beagle
dogs. Neurochem. Res. 28, 11, 1677–82.
[34] Milgram, NW, Siwak-Tapp, CT, Araujo, J, and Head,
E. 2006. Neuroprotective effects of cognitive
enrichment. Ageing Res. Rev. 5, 3, 354–69.
[35] Moe, R, Kingsley-Smith, H, Kittilsen, S, and Bakken,
M. 2004. Anticipatory behavior and emotional
expressions in farmed silver foxes. In Proc. Int. Cong.
Int. Soc.Applied Ethology, 63.
[36] Mueller-Paul, J, Wilkinson, A, Aust, U, Steurer, M, …
and Huber, L. 2014. Touchscreen performance and
knowledge transfer in the red-footed tortoise. Behav.
Processes 106, 187–192.

[37] Müller, CA, Schmitt, K, Barber, ALA, and Huber, L.
2015. Dogs Can Discriminate Emotional Expressions
of Human Faces. Curr. Biol. 25, 5, 601–605.
[38] Neustaedter, C and Golbeck, J. 2013. Exploring pet
video chat. In Proc. Computer supported cooperative
work - CSCW, 1549.
[39] O’Hara, M, Schwing, R, Federspiel, I, Gajdon, GK, and
Huber, L. 2016. Reasoning by exclusion in the kea
(Nestor notabilis). Anim. Cogn. 19, 5, 965–975.
[40] Pickwick, K. How much are Brits spending on pets?
http://www.petbusinessworld.co.uk/news/feed/how-
much-are-brits-spending-on-pets-
[41] Pitteri, E, Mongillo, P, Carnier, P, Marinelli, L, and
Huber, L. 2014. Part-based and configural processing
of owner’s face in dogs. PLoS One 9, 9, 1–11.
[42] Pohl, I. 2011. Quantity Discrimination in Domestic
Dogs. Otto-von-Guericke University Magdeburg.
[43] Range, F, Aust, U, Steurer, M, and Huber, L. 2008.
Visual categorization of natural stimuli by domestic
dogs. Anim. Cogn. 11, 2, 339–47.
[44] Range, F, Heucke, SL, Gruber, C, Konz, A, Huber, L,
and Virányi, Z. 2009. The effect of ostensive cues on
dogs’ performance in a manipulative social learning
task. Appl. Anim. Behav. Sci. 120, 3–4, 170–178.
[45] Riemer, S, Müller, C, Virányi, Z, Huber, L, and Range,
F. 2016. Individual and group level trajectories of
behavioural development in Border collies. Appl. Anim.
Behav. Sci. 180, 78–86.
[46] Rooney, NJ, Bradshaw, JW., and Robinson, IH. 2000.
A comparison of dog–dog and dog–human play
behaviour. Appl. Anim. Behav. Sci. 66, 3, 235–248.
[47] Salvin, H, McGreevy, P, Sachdev, P, and Valenzuela,
M. 2011. Growing old gracefully—Behavioral changes
associated with “successful aging” in the dog. J. Vet.
Behav. Clin. Appl. Res. 6, 6, 313–320.
[48] Savulich, G, Piercy, T, Fox, C, Suckling, J,…and
Sahakian, BJ. 2017. Cognitive training using a novel
memory game on an iPad in patients with amnestic mild
cognitive impairment (aMCI). Int. J.
Neuropsychopharmacol. 20, 624–633.
[49] Seligman, M, Steen, T, Park, N, and Peterson, C. 2005.
Positive psychology progress: Empirical validation of
interventions. Am. Psychol. 60, 5, 410–421.
[50] Soladapt. 2017. Touch Genie Touchscreen Overlay.
https://www.soladapt.com/
[51] Starling, MJ, Branson, N, Thomson, PC, and
McGreevy, PD. 2013. Age, sex and reproductive status
affect boldness in dogs. Vet. J. 197, 868–872.
[52] Stephan, C, Wilkinson, A, and Huber, L. 2012. Have
we met before? Pigeons recognise familiar human
faces. Avian Biol. Res. 5, 2, 75–80.
[53] Steurer, M, Aust, U, and Huber, L. 2012. The Vienna
comparative cognition technology (VCCT): an
innovative operant conditioning system for various
species and experimental procedures. Behav. Res.
Methods 44, 4, 909–18.
[54] Stewart, L, MacLean, EL, Ivy, D, Woods, V, … and
Hare, B. 2015. Citizen Science as a New Tool in Dog
Cognition Research. PLoS One 10, 9, e0135176.
[55] The pet food manufacturers’ association. 2016.
http://www.pfma.org.uk/pet-population-2016
[56] Travain, T, Colombo, E, Grandi, L, Heinzl, E, … and
Valsecchi, P. 2016. How good is this food? A study on
dogs’ emotional responses to a potentially pleasant
event using infrared thermography. Physiol. Behav.
159, 80–87.
[57] Väätäjä, H and Pesonen, E. 2012. Please, Don’t Drive
Me Nuts! Experience Goals for Dogs. In Workshop on
“How to utilize user experience goals in design?” at
NordCHI2012.
[58] Wallis, LJ, Range, F, Müller, CA, Serisier, S, Huber, L,
and Virányi, Z. 2014. Lifespan development of
attentiveness in domestic dogs: drawing parallels with
humans. Front. Psychol. 5, 71.
[59] Wallis, LJ, Szabó, D, Erdélyi-Belle, B, and Kubinyi, E.
The dog’s personality and shared activities predict
owner positive attitude to their dog. In prep.
[60] Wallis, LJ, Virányi, Z, Müller, CA, Serisier, S, Huber,
L, and Range, F. 2016. Aging effects on discrimination
learning, logical reasoning and memory in pet dogs.
Age. 38, 1, 1–18.
[61] Wemelsfelder, F. 2005. Understanding the Tedium of
Confined Lives. In Mental health and well-being in
animals, McMillan, F. Blackwell, Oxford, 79–93.
[62] Wise, RA. 2004. Dopamine, learning and motivation.
Nat. Rev. Neurosci. 5, 6 , 483–494.
[63] Würbel, H. 2001. Ideal homes? Housing effects on
rodent brain and behaviour. Trends Neurosci. 24, 4,
207–211.
[64] Zeagler, C, Gilliland, S, Freil, L, Starner, T, and
Jackson, M. 2014. Going to the dogs: Towards an
interactive touchscreen interface for working dogs. In
Proc.27th ACM, UIST, 497–507.
[65] Zeagler, C, Zuerndorfer, J, Lau, A, Freil, L, …and
Jackson, M. 2016. Canine computer interaction:
towards designing a touchscreen interface for working
dogs. Proc. Third Int. Conf. Anim. Interact. ACI, 1–5.
[66] 2017. http://www.lincolnpetscando.co.uk/gallery.php
[67] 2017. http://doglab.yale.edu/our-research
[68] 2017.https://www.amazon.co.uk/PetSafe-Treat-
Remote-Reward-Trainer/dp/B0010B8CHG
[69] 2017. Pet Tutor®. https://smartanimaltraining.com/