Farmers are under pressure to improve yields and reduce wastage to feed the ever-increasing world population. But they are being asked to do so using sustainable practices, while still maintaining economically viable production methods.
Smart farming using Internet of Things (IoT) technology is seen as one way of meeting these goals. IoT is about providing the right information, at the right time, so the farmer can make the right decisions to get the best from his soils, crops and animals.
IoT in agriculture
IoT applications in agriculture can be divided into five main areas: arable (including fruit and small vegetable farming); livestock; aquaculture (fish farming); equipment monitoring and telemetry; and smart monitoring for compliance with government regulations such as nitrate directives.
For large-scale arable farming, the aim is to improve crop yields by more precise irrigation, application of fertilisers and pesticides, tillage, field preparation, planting, harvesting and storage. Precision farming also means reducing input costs for fuel, fertilisers, pesticides and labour.
Field sensors can monitor solar radiation, temperature, humidity, soil moisture, soil pH and so on to help the farmer better manage the variables in the fields. IoT sensors can be deployed to storage facilities to monitor the condition of grains and vegetables, thereby reducing waste.
IoT applications for livestock are mostly based around tracking and monitoring health and fertility. For example, the Afimilk Silent Herdsman neck-collar monitoring system is used to detect estrus and health problems in dairy cows. The neck-collar continuously monitors the cow’s heat expression and health and transmits the data wirelessly to the farmer’s device.
State of play
There are plenty of examples of these kinds of IoT applications in farming to increase yields and combat disease. But the kind of mass adoption of the technology that would enable the global improvements in yields that are being sought, and the Big Data analysis that would enable predictive farming, have yet to happen.
“I don’t think the issue is anything to do with the farmers themselves, they understand the need to drive efficiency,” says Paul Gudonis, president of Inmarsat’s Enterprise Group. Instead, he points to the cost of completed IoT units, connectivity prices, and the confusion caused by the multiple low-power wide-area (LPWA) network options now on offer, as some of the key barriers.
The choice of connectivity remains a vexed issue. There are a bewildering number of LPWA connectivity solutions to choose from, including Sigfox, LoRa, Ingenu, Weightless, Bluetooth, Wi-Fi, 2G, 3G, 4G and now the new narrowband 4G LTE offerings of LTE-M and NB-IoT. If you are an IoT device manufacturer with global ambitions, which do you choose?
The answer from Spanish firm Libelium, which offers no fewer than 120 different types of sensor, is – the lot. Libelium Waspmote Plug & Play IoT devices support regional variations of 802.15.4 (2.4GHz), 868MHz, 900MHz, 2.4GHz Wi-Fi, 4G and LoRaWAN. And now there are the additional narrowband cellular options of EC-GSM-IoT, LTE-M and NB-IoT to think about as well.
Libelium’s solution is certainly one way of getting around the ‘choice of connectivity’ problem, but it is not one that appeals to many manufacturers or one they can necessarily afford, especially if they are a start-up. It is early days yet for the new technology, but with the encouragement of mobile operator Telia, two Norwegian firms have opted to use NB-IoT.
Irrigation with NB-IoT
Horten-based technology company 7Sense Products believes it has developed the first NB-IoT product for agriculture in the world in the shape of its irrigation sensor. “We focused on irrigation first as it is a real pain for farmers,” says CEO Max J. Tangen.
He explains that irrigation is mostly carried out at night to avoid evaporation and involves a lot of manual labour fitting multiple pipes and connectors together to link the water source and the machines out in the field. Farmers have to go out and reset the machines at the end of each run, so sleep is disrupted.
An irrigation wagon is fitted with big hose reel and a sprinkler at the end. The machine is driven by water pressure, so the reel pulls in the hose. It extends 200-300 metres out in the field and moves about 10-20 metres per hour. It can take 10-20 hours to do a full irrigation run in a field, Tangen explains.
It is very important to maintain the correct water pressure. If it is too high it can damage the crops, too low and the fields are not sufficiently irrigated. “Our product helps them get control of this process,” says Tangen.
“The product is attached out on the sprinkler. The sensors in the IoT device pick up the water pressure. If something happens somewhere along the pipe system and the pressure drops, the device will send an alert to the farmer and he can go out and fix it.”
7Sense opted to try NB-IoT at the instigation of Norwegian mobile operator Telia, which was looking for partners to help trial the new standard, including chipset-maker
u-blox and 4G and NB-IoT cellular infrastructure provider Huawei.
“We were attracted to NB-IoT because it is a general open standard,” says Tangen. “Then it is a question of deployment. Telia should have Norway covered by October 2018 for NB-IoT. Trials show that NB-IoT delivers on the promises and it really suits our application. The farmers who have used it love it.
“MNOs are looking for enablers to drive traffic, and for us it has opened up new ways to market we didn’t anticipate. Before we went through traditional distribution channels, but now we may get traction through Telia and Telenor and they can take our products to global markets.”
Shiip tracking
The other Norwegian firm, Shiip, along with co-located development company Nortrace, has also worked with Telia, u-blox and Huawei to trial its tracking device for sheep. “Our ambition is to use the technology to better the welfare of the sheep and to make life easier for the farmer,” explains Eric Macody Lund, Shiip’s director of business development.
The sheep graze freely for approximately 200 days and then have to be found and brought back to the farm for both welfare and regulatory reasons. The traditional way of herding sheep is by sending five to 10 people, and possibly a helicopter, to search for and herd the sheep across very large areas. “It is not cost- or time-efficient and it involves a lot of manual labour and it is tough on the shepherds,” says Lund.
The Shiip product combines NB-IoT with GPS so farmers can pinpoint where the sheep are. “You can also create geofences to see if the sheep have left a designated area, so NB-IoT technology suits this user case very well,” notes Lund.
The tracking product is based on NB-IoT chips from u-blox with a smaller Huawei chip inside that. In 2017, the tracker was fitted to 1,000 sheep in Norway, with Huawei and Telia very hands-on in helping to optimise the network and troubleshoot on the fly.
“The key finding was that it worked,” reports Lund. “The battery life was fine and the connectivity was very good in good-coverage areas. Some farmers were disappointed in areas where the connectivity was not so good. But overall it has helped steer our direction and identify areas that need improvement.”
Coverage is a problem, as MNOs are notorious for not providing adequate coverage in sparsely populated rural areas. Lund agrees that MNOs may have to add more rural base stations, but suggests there may be good economic incentives to do so.
“We are trying to build a case for MNO investment in rural areas. The human population may be small, but if we add 20,000 NB-IoT connections around a base station, does it then make economic sense for the MNO to install a new base station?” asks Lund.
“One of the reasons Telia is involved with us is they want us to help them understand the user cases and the economics of deploying more coverage. 4G coverage in urban areas is 100 per cent, but maybe if we help increase the connection volumes with IoT, that will encourage more rural base station deployments,” he observes.
Big Data vision
UK-based start-up Pycno has developed an arable multi-sensor that measures solar radiation to record the amount of sunlight, air temperature and humidity, soil temperature and soil moisture every 15cm down to a depth of 1.2 metres.
The company has two sensor products. The first gets around the ‘which connectivity solution to use’ issue by integrating a special SIM card that works in more than 160 countries. It is not locked to any particular operator. Instead the SIM just locks into the best network for the area and transmits data directly.
The second product works by creating a mesh between clusters of sensors, which send data to a master sensor, which in turn sends the data to the cloud. “Right now they can connect using Bluetooth, LoRaWAN or Wi-Fi if they have it,” says co-founder Nahuel Lavino.
“We are looking at Sigfox and NB-IoT. We have control of the whole supply chain, so we can easily add NB-IoT.” Pycno has not used satellites so far, but Lavino says: “Satellites will make a huge change especially for sensors out in the field in areas where there is patchy or no terrestrial coverage.”
For Lavino, one of the key barriers hindering wider adoption of IoT in agriculture is distribution. “Our main challenge is finding the right kind of distribution. Whoever sells equipment and machinery to the agriculture sector is not selling IoT products at the moment. We need more education around the farm machinery distribution channels. The channel needs to let the customers know what is on the table.”
Education among farmers is another barrier to IoT adoption. “Farmers are always looking for new technology,” concedes Lavino, “but they are also very sceptical, so it is hard to get traction with them. They say: ‘Wait a year until we’ve seen how it works with this year’s harvest.’
“The problem is not about not understanding the technology, because we made the installation so simple. It takes one minute to set up. The main problem is getting the farmers to understand what the data means. That takes time.”
For this reason, the company has put a lot of effort into its platform and analytics tools. “If you turn the raw data into a heat map, the farmers can see what the data is telling them very simply in real time. You need more water here and less there. Why? Because of the topology of the soil and rocks underneath.”
Pycno’s real goal, however, is not the hardware but a focus on the Big Data opportunities for IoT in agriculture. “For us to accurately predict agricultural conditions better everywhere, for every crop, we need loads of data. Even a bit of data will help an individual farmer, but the more data we have, the better the analysis and the more accurate the predications we can provide, which will help everyone,” argues Lavino.
End-to-end solutions
Farming is a tough business and many farmers already face cost pressures, so adding IoT into the mix can seem too difficult and too expensive. Inmarsat’s Paul Gudonis argues that the industry needs to put together solutions involving all the necessary partners in the IoT stack, so farmers can sign up to a ready-made, end-to-end solution.
“We are looking to develop strategic partnerships with a small number of organisations to provide delivery-specific solutions into the marketplace. There needs to be a very clearly focused scheme that really shows what the technology can do,” argues Gudonis.
Gudonis admits the cost of satellite connectivity has precluded its use to some extent. “Satellite costs have definitely been an economic barrier in the past, but we think that will rapidly disappear,” he says.
He argues that satellite is particularly useful in areas with poor or no terrestrial communications, but he reports that even regions with good cellular coverage are interested in the possibilities of satellite. “If you are relying on data to run a business, you want to know that service is going to be reliable,” points out Gudonis.
“Satellite provides very strong levels of communication. The reasons mariners use us for safety communications is the same selling point for farmers and industrial IoT connectivity solutions – it’s reliable.”
Inmarsat is working on various proofs of concept (PoC) trials using its Inmarsat-4 Broadband Global Area Network (BGAN), working in conjunction with LoRa technology to collect data in the field. The data can then either be sent directly via a BGAN terminal on each LoRa access point, or multiple sensors send data to a LoRa gateway, which is then backhauled to a cloud-based network server by satellite.
Among other PoCs, Inmarsat is trialling plantation monitoring in Malaysia, Indonesia and Peru using BGAN M2M and LoRa-enabled weather stations, soil monitoring sensors, an asset tracking solution and a smart IoT platform for the client. The aim is to help the farmer know the optimum time to harvest particular areas of the plantation, to improve operational efficiency and to save on fertiliser inputs.
But Gudonis also believes a wider awareness of what IoT and satellite-based solutions can provide is needed. “We need to get the first few success stories out there. We hope to have some PoCs brought to conclusion in the next few months and then we need to show how our solutions can drive efficiencies and improve yields.
“When IoT first started we had multiple sensors generating lots of data, but in a form that was hard for the customer to use. Now sensors are designed around solving a problem, rather than just gathering the raw data,” observes Gudonis.
Clearly, there is plenty of activity in the IoT in agriculture sector and no shortage of ideas and options. But unit and connectivity costs do need to come down. It is also too early to know whether any LPWA connectivity standard will emerge as the dominant technology, so device manufacturers still face difficult choices.
With MNO backing, NB-IoT should certainly take off, but probably several standards will remain in use as it becomes more apparent which technologies work best for particular agricultural IoT applications and different geographies.
But there seems to be a considerable degree of optimism that the sector is advancing and the goal of increasing yields in a sustainable way to feed the world’s growing population might be met.
“I think that we are at the cusp of being able to scale up IoT adoption in agriculture very quickly,” says Gudonis.
