Robotics in agriculture – the opportunities and constraints


THE most recent meeting of the Northern Ireland Branch of I Agr E featured a presentation by Dr Debbie McConnell about the app-lication of robotics to livestock husbandry systems.

The meeting was at AFBI, Hillsborough, where Dr McConnell is carrying out related research work on the subject of “Digital dairying – increasing the value of precision technologies in the UK dairy sector”.

She also recently completed a 16-week world study tour funded by the Nuffield Farming Scholarship Trust to the US, Middle East, India, Australia and New Zealand.

The Nuffield Foundation was set up in 1947 to encourage travel to other countries, observe related developments and exchange know-ledge/experience of the science/practices of agriculture.

It co-ordinates funding from various sources including, as in this case, the Thomas Henry Foundation. It was set up in 1847 by the terms of the will of Thomas Henry, Downpatrick, County Down, to support post-graduate science-related studies for agriculture.

World overview of agriculture

Dr McConnell, in introduction, summarised the current role of agricultural production against the background of increasing world human population. Eleven per cent of the world landmass is covered by crops. 30.7 per cent of the world’s population of around 3.36 billion people still work in agriculture.

It is estimated that about third of the world food production is still wasted due to weaknesses in storage, processing or distribution infrastructure systems.

In India, 70 per cent of the food is bartered and in the region of Hydrobad is not distributed beyond a 15 mile radius. Indian agriculture is large scale with the feeding of 133m animals and production of 1.46 billion litres of milk each year.

With the increasing world pop-

ulation, the average area of agri-cultural land per person is just 50 per cent of what it was 10 years ago. Water availability and management is a critical limiting factor across the wide range of climatic conditions.

The example at Quatar was quoted with a 21-day growing season in a warm climate using 30 million litres of recycled waste water to produce one tonne of dry matter.

In the Punjab area of India, where just five per cent of the land produces 40 per cent of the agricultural output, the continuing use of irrigation from ground water has lowered the water table by one metre/year.

By contrast in high density urban economies, such as Singapore, high-value horticultural crops are now being grown intensively in high-rise controlled environment buildings.

Use of digital technology in agriculture

The 700 per cent increase in internet data use has revolutionised communications and the storage of large amounts of information. There is now a strong bias to the use of mobile and wearable communication devices. The digital collection and use of data is helping to maximise the utilisation of labour, feed and land resources for livestock enterprises. Examples in the dairy industry include the control of robotic milkers and monitoring of herd activity for heat detection.

Grassland management

Aerial image pictures taken from a UAV or drone are now being used to monitor and manage arable field crops. Using the same technology over grassland offers the potential for rapid collection of useful management data. The analysis of aspects of the images, not all visible to the naked eye, may be able to quantify herbage yield (as a convenient alternative to the existing falling plate or trailed plate methods).

It may also develop to assess protein content, ME and DM yield for a standing grass crop. This information can be co-ordinated on electronic field maps with other information such as soil nutrient sample results. In favourable conditions, a drone can record pictures at around 500 acres/hour. Current limitations include:

q A high definition camera is a relatively heavy load for some drones.

q Good visibility conditions are needed to achieve clear pictures.

q The taking of 2D pictures above sloping land can under-estimate the amount of surface area.

q Their use, in terms of height and range, may be restricted by public security and air traffic restrictions.

q Smaller drones may even be attacked by birds of prey!

Monitoring animal condition and behaviour

A variety of animal wearable electronic devices are already available to measure activity. A pedometer will record movement patterns to detect heat behaviour, lameness and rest/sleep patterns. On-body devices record heart condition, body temperature and blood pressure.

One of the latest is an electronic data logging halter, with a pressure sensor in the nose band, to record grazing and rumination activity throughout the day. It can be worn in the field by an animal for up to 60 days without attention. Improved electronics and build quality mean that reasonably priced, more durable, commercial versions are now available.

Electronic herding and control

In Australia and New Zealand some work has been taking place with robotic programmed herding devices to control cow movement between paddocks.

The concept of virtual fencing using electronic collars (with various frequency sound emissions or mild electric voltage pulses) to steer individual animals to preferred grazing areas, without physical fencing, is also being successfully tested.

This system can also segregate individual animals.

An electronic calving alert uses an attached movement sensor (suitable for field or indoor use) to measures contractions and record how long a cow has been in labour. If this exceeds a specified time it alerts the herd manager to the need for assistance.

Robotic milking

Robot milking units are now well established and available as single or multiple units. Pictures from a 24 unit rotary parlour in Tasmania were shown of its automatic pre-washing, cluster attachment, removal and teat dipping. Cows use the parlour an average of 2.6 times/day.

An example of remote herd management in New Zealand is a 400 cow spring calving unit with employee operators’ activity managed electronically on screen by the owner in his farm office 300km away!

Who is buying the technology?

It is more attractive to a younger generation, with more training and confidence in the technology, seeing it as a way to reduce labour requirement and improve management control. Experienced stockmen may see it as interesting, and in some cases highly desirable, but not ready to take over from the direct application of their time-proven stock person skills and knowledge.

Robotic milking provides the attractive option of getting a break from the seven days per week routine. A survey of UK dairy farm personnel reports that 45 per cent felt that they would need better ICT skills to make best use of the available technology.

The typical time scale for how people adapt to the full use of new information systems is three to six months for initial familiarisation, up to 12 months to consider it as a fully accepted part of their daily work routine and longer for wanting to learn more and seek its further development.

Cost benefit analysis needs linked evidence base

As the technology is so new it is still difficult to assemble enough test data to truly quantify the benefits for its widespread adoption. An International AMS (Automatic Milk-ing Systems) Research project is now under way, including Dr McConnell’s research work at AFBI, to co-ordinate and analyse the wide range of available data.

It should be possible, using this information, to improve herd performance and welfare as well as making life easier for stock people. We wish her every success.

A wide ranging technical discussion followed before the chairman thanked Dr McConnell for her most interesting and informative presentation.

Further I Agr E evening meetings are arranged for Gilfresh, Loughgall (February 13) and KDM hire, Loughry Campus (March 1). As always, visitors will be welcome. Contact for further details.


Please enter your comment!
Please enter your name here