Will Robots Take Over Farms? How Smart Sensors Are Revolutionizing Farming Today

March 29, 2017

By Dr Ng Sing Muk

Farming is often associated with hard labour and that one will get dirty and muddy. It is unusual for youngsters these days to set their ambition as farmers. Being a farmer isn’t a glamourous job, and the reality is worsened whereby the farming sector faces lots of challenges such as uncertain weather patterns, frequent disease outbreaks, pollution, increasing energy costs, high costs in maintaining soil fertility as well as stiff competition of land use from other urbanisation activities.

The future seems bleak but we have no choice but to sustain this sector as food is essential to life. The world population is growing fast hence the pressure to produce enough food is increasing by leaps and bounds too. Just imagine, the amount of food required in a day to feed the world’s population of over seven billion people to date, compared to just half this number in the seventies. The population is expected to reach about 10 billion by the year 2050. By then, assuming each person is consuming two bowls of rice on average, putting the bowls together side by side can make a straight line that takes us to the moon and back at least four times!

It’s good news then that the industry is responding quickly to prevent this global challenge from growing into a crisis.  One of the ways to solve this problem is to switch from conventional farming to smart farming. This model employs a more precise and efficient way of farming techniques using technology. The environment of the farms is set to be well controlled and regulated to operate at its optimum condition. This means ideally the application of water, fertilizer, pesticides, and light will be made only when required by the crops but not based on routine schedule determined by time scale.

The amount applied is fixed by consumer demand and harvesting is also performed at the correct timing to ensure the best quality of crops. As such, crop yields can be improved significantly while solving some existing challenges especially on wastages of resources and energy.  The concept can be convincing but the reality is far more dynamic and complex to achieve. We are dealing with crops and its surrounding that will not be able to communicate directly with us and how can we respond accordingly if we have no idea on the needs of the crops?

The core task is to translate the real conditions of the farm and the information of the crop into some readable signals, where interpretation can be made and subsequently lead to required actions. Some might start wondering on how this is made possible, and the answer is simple – the utilisation of sensor technology.

In the context of smart farming, these sensors will act as the ears, eyes, nose, skin, and tongue to the farmers. They are the first channel to collect continuous information of the farm before being interpreted for further action. The monitoring of physical parameters such as temperature, humidity and light intensity in the farm is quite straight forward as these physical sensors are well established and can be easily obtained from the market. There are even sensors that offer plug-and-play features, where the sensors can operate directly once connected to a computer using the USB port.

Colour changes picked up by these physical sensors can also be used as an indicator to determine the maturity of a crop; for instance, the ripeness of fruits. However, it is more complicated when it comes to the detection of chemical and biological parameters for example mineral content, crops chemical composition, pesticides level, disease risks, water quality, microbes and nutrient levels in soils that are crucial to the farming industry.

Usually, chemicals or biosensors will be used but not all are commercially available. Thus, modification and optimisation of the sensors might be required, or else, one might need to develop their own sensor. This is not an impossible mission, and what is required is just a good understanding of the basic knowledge governing the working principles of a sensor.

Generally, a sensor is a simplified device that can translate physical, biological or chemical properties into a readable output. A simple sensor consists of a sensing receptor, transducer, detector, data analyser, and output readout. The basic working concept is similar to the human senses that we have. The sensing receptors are at the outermost interface that will get in contact with the target of interest such as fertilizer content in soils or pesticides level in crops. Sensing receptors have the ability to attract and interact specifically with the target, but ignoring the rest of the other non-relevant species. It can be made of synthetic materials such as polymers, or even biological species such as antibodies.

The interaction with the target can cause the change in the transducer property such as the vibration mode, light wavelength or electrical resistance. A detector is used to pick-up this changing properties and the degree of change will be correlated to the actual concentration of the target species by a data analyser. Once this is done, the final output readout will be presented in the form of digitised values that make sense to us.

Having sensors in place, the operation of the farm can simply tap on the existing technology on the Internet of Things or commonly known as IoT. Information from various sensors can be integrated and sent to the control room via the Internet and similarly, actions can be activated from the control room and sent back to the farm site for implementation by machines or automated devices through the same route.  Bear in mind that the control room can be miles away or even located in the city, while the farmers can operate their farm in suit without getting themselves dirty. This might turn the impression of a farmer into a cool profession and setting job opportunities for the youngsters who most of them are very familiar in using electronic gadgets and information technology. 

Smart farming is the future and it is only a matter of time before it becomes mainstream. Crops producing countries should take proactive steps to ensure a smooth transition and start implementing the initiative in stages. It might not be too difficult and could start just from the core – getting and developing the right sensors on the farms.

Dr Ng Sing Muk is Associate Director of Research and Consultancy Office, at Swinburne University of Technology Sarawak Campus. His research interest is in the area of chemical and biosensors for various applications. He is contactable at smng@swinburne.edu.my