With the population growing rapidly
the demand for food is increasingly high. Farmers are decreasing in numbers and
farms growing in size, consequently there is less time to monitor each animal.
Concerns are also growing for animal health with respect to food safety as more
animals being reared in less space with less resources (Dawkins 2014). Farms becoming more intensive is
often seen to decrease welfare standards of these animals but often there is an
overlap in increasing efficiency and increasing welfare. For example, by
improving litter under chickens feet that gives off less ammonia will improves
ventilation and therefore air quality which will reduce mortality and damage to
birds (Dawkins et al. 2004).
Precision Farming is the new
revolution caused by the development of new technologies that increase
efficiency of production. This ‘per animal’ approach means each individual is
monitored in aid to reduce mortality, waste and improve their living
conditions. These new technologies like sensors, automatic surveillance,
wireless communication tools, internet and cloud storage allow farmers to
monitor their livestock more efficiently. Farmers can check their animals when
they aren’t present and are able to catch lameness in early stages to take
pre-emptive measures. As well as nutritional requirements can be worked out
more effectively so animals welfare is promoted.
Automatic Milking Systems
Automatic milking systems were one of
the first precision farming developments. The most obvious welfare benefit of
these systems, is that it allows cattle to choose when they are milked. Since
the cows are milked voluntarily it reduces the need for human handling and
cattle milked in a AMS have shown to have a lower heart rate and lower
adrenaline levels than those in a tandem milking parlour (Hopster et al. 2002).
Most AMS promote udder health due to their automatic teat cleaning and milking
cup-attachment process. Their interfaces are designed to minimize the chances of
intramammary infection caused by milking and prevention of diseases such a
AMS require cows to be milked
independently of their herdmates in single stalls which goes against cow
behaviour and therefore such social isolation may affect welfare. Lame cattle
have also been shown to visit AMS systems less often, low milking frequency can
indicate this issue, cows may be hungry but refrain from visiting the AMS. Some
strategies to encourage cattle to return from pasture to be milked can cause welfare
concern, such as limiting water supply, can reduce water intake which in turn
decreases animal’s welfare (Jacobs et al. 2012).
Although the most common way of
encouraging cattle to be milked is by the feeding of concentrates. Automated
feeders controlled by computers can recognise individual animals that
automatically record animal’s feeding behaviour. As data from these
technologies can not only give each animal their optimal quantity of feed for
efficient production but also quickly identify a change in eating patterns and
detect early signs of disease. In the ‘Five Freedoms’, freedom from hunger is
the least controversial and technologies like these could be valuable in
allowing automated assessment of hunger (Rushen et al. 2012). Rewarded automatic feeders can also be important in
the weaning of calves and identifying the animals who are struggling to
adapt. As more animals are housed in bigger groups means lameness is harder to
detect, the automation of feeders are a useful tool in prevention of these
diseases. One problem with this is that equipment on farms is different making
this assessment of animal welfare less useful at a group level.
Many devices are attached to animals
to monitor their behaviour and movement, most common are accelerometers. The
time an animal stays standing or lying down is important in measuring its
welfare, for example a short length of time spent lying down can suggest stalls
are not suitable (Rushen et al 2008) or an excessively long time spent lying
down suggests there is a high level of lameness. Similarly tilt switch activated
technology are also small devices usually on the leg which measure when the leg
is horizontal so less time is spent checking the herd. Accelerometers can also
measure the number of steps made by animals and can use automated gait scoring for
One benefit of devices attached to
animals is the group welfare assessment and being able to be transported
easily. They can also be used on free roaming animals too which otherwise data
is hard to collect. The main issue with these accelerometers is that they are
invasive to the animal and have potential to influence the animal’s natural
behaviour and in rare circumstances harm them. Having these devices on animals
also increases the amount of handling time to attach/remove them which is
likely to cause the animals more stress (Rushen et al 2012).
Accelerometers are more accurate when
on the leg rather than the neck of an animal and rely on the data being
‘cleaned’ especially when they take many samples in a second. Due to the memory
of the accelerometers sampling intervals must be made practical but also
measure behaviour accurately (Ledgerwood et al. 2010). Accelerometers assume
the animal is lying down when their leg is horizontal but can also occur when
grooming etc. This could also be mistaken as the animal being uncomfortable and
restless. Data also requires editing in analysis, like the inter-visit
intervals in feeding which could be explained by loss of signal which in turn
can cause a misjudgement of welfare levels.
Biosensors are devices that quantify
biological components of livestock including temperature, pressure, movement
and glucose (Neethirajan et al.
2017). With climate being one of the major limiting factors of efficient
production, increases in temperature reduces performance and cause an increase
in mortality. This causes concern for welfare, so temperature sensors have an
economic benefit as well as being advantageous for animal health. Along with
pH, temperature can be used for calving alert and rumen functionality. The
development of nanobiosensors, monitoring of welfare is made less invasive
promoted as well as ethical handling of animals being promoted. Nanotechnology
focuses on quantifying stress and metabolic biomarkers based on animals activity
for example oestrus monitoring and detecting lameness (Neethirajan et al. 2017).
Although biosensors have huge
potential in improving animal welfare, there is still issues with many farmers
having the skills to utilize these technologies effectively (Van Hertem et al. 2017). As
analysing the data from sensors correctly takes a lot of time as well as many
different formats and frequencies, there is more room for error in interpreting
the results which could put animals’ welfare at risk for example monitoring
metabolic biomarkers could mean a stockman may miss the early stages of
disease. Farmers require proper training to interpret these results and they
should aid real-time monitoring.
Increasingly smartphones are being used in
livestock monitoring. In poultry farming, smartphone cameras have been used to
monitor flock movement which can detect unusual patterns to determine if the
flock is mainly healthy with birds moving around or not. It also allows farmers
to predict which flocks will be most efficient or which will have the highest
mortality and leg damage development (Dawkins 2014).
detections like this, pre-emptive measure can be taken, and it allows the
reduction in unnecessary antibiotic and medication use.
Image and Sound Analysis
Developments in image and sound analysis mean
programmes can ‘read’ these images and determine animal behaviour as well as
automation to recognize certain different sounds. Many different uses of these
programmes are being explored for example measuring space between cattle
standing and lying down to estimate stall sizes (Ceballos et al 2004) or
tracking activity levels of animals and any features like an arched back in
relation to lameness. The limitation of this is the difficulty of seeing
individual animals in large herd. A different approach is thermal imaging at
group level to assess temperature adequacy of pens. Classification is important
in more vocal animals like pigs, simple measures of amplitude of sound can give
pigs responses to the likes of temperature and humidity of housing as well as
detection of coughing by sick animals (Rushen et al. 2012). The main benefit of this technique is it
is non-invasive and wont effect the animals natural behaviours.
Issues with automated monitoring concern the
validity of the data they produce, if they are measuring the correct thing. As
well as the accuracy of these technologies, if there the right sensitivity or
specificity. The best method of making sure the monitoring is valid is to
compare it to human observation. Although it is assumed that automation is more
accurate than human observation, it is not always the case as the electronic
signals between operating computers for data storage and control unit on the
farm could effect the accuracy for example automatic feeders not recognizing that
a cow is present.
Other technologies are less effective, ground
reaction force has problem with low sensitivity of detecting lameness and
ineffective with gait scoring but development of pressure-sensitive walkways
have proved more effective (Maertens et al 2011). Detecting lameness is much
more effective when data from the likes of weight distribution, walking speed
and lying time in cows, accuracy raises by 20% than that of one set of data
alone (Chapinal et al 2010).
A major concern of precision farming in animal
welfare is that there will be a focus on only the behaviours that are measured
by computers and missing out on real observation, we may know animals are
standing up but not what they’re doing for example if they are fighting
(Rushenrtal et al 2012). Another issue being the reduced human and animal
interaction time, automation allows a huge amount of information to farmers
about their livestock but may mean less time is spent watching their animals.
Precision livestock farming has huge potential
to allow farms to become more intensive whilst keeping up welfare standards.
These new technologies not only increase productivity but also allowing early
detection of diseases, reduced mortality rates, better housing, more efficient
feeding and increased monitoring without human interaction. Despite this, these
computer systems need further development to become completely accurate and
valid as well as better skills in interpretation and analysis of this data is required
to utilize the technologies efficiently.