Maturing sensor technologies and the greater accessibility of data techniques like visualization and machine learning are driving a new generation of precision agriculture and computerized farming systems. But whether these will be taken up by industry at scale will depend on answering one key behind-the-scenes question: does the new tech have an API strategy?

The rise of Precision Agriculture

Sensors are becoming cheaper, more robust and are now able to measure more environmental factors necessary to encourage sustainable farming practices (like soil water composition, and pest risks). Alongside this, new software is enabling that sensor data to be collected in real-time, analyzed for crucial insights, and visualized in an easily understandable way. The data is also beginning to be added to machine learning, such as instant-alerts that identify when particular environmental thresholds are reached. Now, several of these technologies are adding automation techniques to allow the software and sensors to respond independently, for example, by automatically watering crops when soil moisture falls below a certain level.

The difficulty is that as an emerging industry, precision agriculture is seeing a range of sensor makers and software providers enter the market. Some of those are large scale hardware manufacturers, some are startups, and some are industry-specific providers who are trying to diversify their current products. Many try to offer a software front-end to their sensors so that farmers can read and collect data from the sensors they offer, but farmers often need something more flexible. They need a solution that can incorporate data from a range of different sensors they may be using from different manufacturers, may already have an existing farm management software they want to keep, or want to choose their own software that has a particular capability that meets their needs.

This is where the API — application programming interface — comes in. APIs are an agreed way of representing the data from one source so that it can be integrated with another. In the same way that a powerpoint doesn't need to know what type of appliance you are plugging into it in order to feed the electricity to it, an API creates a standard way to plug in software so that it can feed the data source to the software.

Making Sustainable Farming and Urban Management a Reality

Projects like Weenat, IrrigNET, and SmartSilo are all making sustainable farming practices more accessible and able to be managed remotely. With these sorts of products, it is increasingly feasible to manage rural farming from an urban base. But for these products to scale and build enough of a market, they need to be able to integrate with a large number of sensors and software solutions.

Emerging French-based startup Weenat, for example, uses durable multi-sensors that can withstand water and tough weather conditions to continue relaying readings. President Jérôme Leroy told an APIdays conference in December last year that poor farming practices are responsible for some of the world's greatest sustainability burdens. For example, 70% of the world's water usage is for agriculture. Leroy believes that technology can improve current agricultural practices and reduce water requirements. He says that one of the ways to leverage a technological advantage is to help farmers implement precision agriculture practices: "We are moving to a trend in agriculture where the farmer wants to get data exactly where they think there will be the problem, not five or eight kilometres away," Leroy said.

Leroy's vision is shared by other projects like irrigNET, a project from Serbian startup DunavNet. IrrigNET sees one of the key challenges to successful crop production is rational irrigation, and the ability to understand and respond to plant and soil characteristics in realtime. Using sensors that measure soil moisture at various depths and air temperature, irrigNET can not only assess current environmental factors, but can also create automated 'recipes' that trigger more efficient farming practices. For example, farmers can factor in expected rainfall and air temperature, and trigger irrigation watering automatically when soil moisture falls below a certain threshold. While systems can be set up automatically, realtime monitoring dashboards still give farmers control if they want to reset values at any time.

Meanwhile, Spanish startup Ubikwa Systems is building SmartSilo, a sensor-based system to monitor and optimize stock in the silos of livestock farms. More sustainable stock management can reduce CO2 emissions by 25%, and in Europe alone can save farms more than €2 billion a year, according to CEO, Jaume Gelada.

Integrating Farming Tech with APIs

At Weenat's conference presentation in December, Leroy confirmed the next major task on the startup's roadmap is to build an API. While Weenat comes with it's own dashboard software for monitoring the multi-sensors used in the field, Leroy acknowledges that farmers may be combining that data with other sensors they are using, or already be using a farm management software solution. An API would allow Weenat's farmers to integrate the sensor data into their current systems.

Both irrigNET and Ubikwa have chosen a different route: they use the open source API platform FIWARE to make their sensor data available for integration into a farmer's preferred software system.

"Our goal was to develop a solution that is affordable, easy to install, available for all farms and that provided all the apps and services needed to optimize the logistics associated with animal feed stocks," says Ubikwa's Gelada. "So there were a lot of requirements we needed to take into account: We needed standardization to achieve maximum interoperability, we wanted to be open source to avoid technology that locks users in to proprietary systems and limits our growth capability, and we wanted an engaged community that helps drive a continuous stream of innovation."

Gelada says the team looked at FIWARE as a possible way to build the API for their sensor, but were hesitant at first. "When we began our research, FIWARE was a very ambitious project launched by the European Commission. It looked interesting and we decided to make some tests and assess how well they were performing their mission. Even though it was not easy at that time diving into the technology specifications, our first impressions were quite positive. FIWARE was covering most of our needs and was performing well. However, we still had doubts because of the immaturity and uncertain future of the platform. Would it have enough momentum to consolidate by itself after the support and funds received by the European Commission had finished? When the EU announced the launching of an accelerator phase of the program with €150 million to fund 800 startups, we had no doubt that FIWARE was going to play a key role in the market."

IrrigNet was one of those startups funded with some of that available funding from the accelerator phase. They used their startup funding to pull data from sensors alongside realtime data from third party sources like weather data to create an optimized and automated irrigation system that increases yield, reduces water waste and prevents land erosion.

Jose Manuel Cantera, a Tech Expert with Spanish telco Telefonica, which is one of the partners in the FIWARE initiative, urges startups to avoid building their own APIs for their new sustainable farming products, and instead work on the open source platform.

"I would say that creating their own platform and APIs could limit further opportunities of integrating their platform in multiple and unforeseen scenarios," Cantera says. "For instance, let's consider a smart water solution that could be developed to optimize water supplies to farms. If a farm has a soil sensing system built on proprietary technologies it will not be able to interoperate, at a data level, with the smart water solution and as a result the farm will be isolated, and will have to spend additional time and money to connect to the water management solution. The main point here is that vertical, proprietary solutions might work initially but might have scalability problems in the end affecting both the end customer and the solution provider."

FIWARE is not the only player offering to manage APIs for sensor hardware manufacturers and for farm management software providers. DreamFactory is another open source tool that helps users create their own APIs more quickly than they could on their own. That differs to FIWARE's model which encourages others to use their common set of APIs.

Precision agriculture has a huge potential in helping reduce farming inefficiencies and creating a more sustainable industry. But it can also facilitate a new relationship between cities and their urban food bowls. APIs, while working behind the scenes to connect various sensors and data systems, will be the true enabler that makes all of these possibilities a reality.