Those born before the luxury of pre-set radio channels will remember the crackling sound of wavelength interference as they tuned their sets to their favourite station. It’s the same with images taken from space. The objective may be to look at something specific on the ground, but all sorts of “interference” from clouds to pollution can distort the picture.
Being able to use drones and other sensing equipment to remotely monitor crops has been a big boon for agriculture. However, visual “interference” remains an issue, and the poor quality and inconsistency of the images coming back is preventing wider adoption of the technology.
About to solve this problem, however, is agtech startup ProvEye and its founders Dr Jerome O'Connell (CTO) and Prof Nick Holden (R&D director), who have developed an automatic image correction system that filters out "interference" to give a much clearer view of what is below.
Their technology is significant because the data from images subjected to ProvEye’s processing software is vastly superior to what’s currently out there. This is crucial because more reliable data means farmers can make better informed decisions about how to improve their yields and reduce their input costs. Specifically, being able to accurately monitor their land from above puts farmers ahead of the curve when it comes to key areas such pest control and crop health and quality.
Tim Buckley is ProvEye's recently recruited chief executive, and he says the company's founders have "solved the technical challenge of taking the 'noisy' optical information collected by drones, vehicle-mounted sensors and satellites and extracting a clean signal from it".
“This then forms the basis for the quantitative modelling and prediction that is very important in large scale farming. We’re talking about users in countries where farms are the size of Kerry, and it would be physically impossible to walk the land,” says Buckley.
“The use of wide-area image services is a huge market opportunity, but its value is not being realised because the data being collected is unreliable,” he adds.
“Our potential customers are businesses selling inputs, agronomic services, precision agriculture services and decision support to farmers, and they are only as good as the data they collect. For example, as of now if the weather changes from sunshine to rain then the data also changes even though the crop stays the same. The ProvEye sensing solution works its magic to correct images for consistency which means that if a change is detected then something has happened to the crop.”
Dr O'Connell is an expert in image processing tools and software for remote sensing, and has worked with a range of agencies including Nasa, the World Bank, Gorta, and the European Space Agency. Nick Holden is professor of biosystems engineering at UCD, and has a background in optical and remote sensing and precision agriculture.
"Jerome and Nick came at the idea for ProvEye at about the same time but from two different perspectives," says Buckley, who is the former head of the digital agriculture division at Origin Enterprises.
“Jerome was doing some work in the UK with drones in sustainable agriculture, and found that there was no consistency in the signal. Nick was researching UAV (unmanned aerial vehicle) use in precision agriculture, and noted the lack of quantitative data applications being developed. Originally the system was designed for satellites – and this is a market segment we will be returning to as it represents a significant opportunity – but for now our technology is being applied to drones.”
ProvEye’s image processing software can be used with any camera in any satellite, on any drone or in any UAV. The camera belongs to someone else, typically the satellite or drone owner/operator, and the ProvEye software is bought as a bolt on. Large-scale users (satellites) will pay per hectare. Smaller users (drones/infield machinery) will pay an annual charge to access the software.
Buckley says ProvEye addresses two main end-users: farmers using digital technology to better manage their land and agri-service providers keen to offer more reliable advice to their clients.
“For farmers, precision agriculture can offer gains of up to €118 per hectare or losses of €141 per hectare if it’s based on inaccurate or poor quality data,” he says. “ProvEye is automatable, thus any customer can insert it into their data analysis pipeline or indeed make it the whole pipeline. It is accurate and reproducible which we know other systems are not. It is sensor agnostic and will work with any unit currently on the market or likely to be developed in the next decade. In short, our technology can be combined with predictive models and embedded in an end-to-end solution for the quantitative prediction of agricultural behaviour.”
ProvEye is a spinout from UCD, and is based at the college's NOVA innovation hub. The company, which employs four, was set up in March, and Buckley says employment will grow to roughly 39 over the next four years. Investment in the business to date is around €1 million between personal funds and support from Enterprise Ireland. The company's first product will be on the market by early 2020.
Over the next five years ProvEye will add a number of new products designed to take the decision support process to the next level by using AI and machine learning to offer deeper and deeper analysis. To build out this pipeline the company is now looking to raise around €1 million in investment.
“ProvEye technology has the scope to become an industry standard. We know that compared to other software ProvEye is more than twice as consistent and at least four times more accurate for crops such as grass,” Buckley says.
To develop its B2B business rapidly and at-scale ProvEye has established relationships with three large global agtech businesses.
“The first is a supplier of livestock farm management software used by over 3,500 farming enterprises. The second is a leading global farm management platform supplier that enables growers and their service providers [agronomists, banks, insurers, accountants] to centralise and standardise data to manage risk. And the third is a US-based but globally active agricultural machinery company supplying everything from sprayers to combine harvesters,” says Buckley.
ProvEye’s final ace is its capacity to contribute to sustainable farming and food production.
“Because we have opened the critical bottleneck that was limiting the use of wide-area optical imaging we are facilitating the optimisation of environmental resource use and the economic value of farming,” says Buckley. “To suggest that ProvEye technology will transform the food system is a step too far, but we know it will undoubtedly contribute to the sustainability of the food supply.”