From left: Prof Nick Holden and Dr Jerome O'Connell. Image: ProvEye
Our Start-up of the Week is ProvEye, a start-up that has developed technology that removes noise from image data to create actionable management decisions in agriculture and other industries.
ProvEye is a University College Dublin (UCD) spin-out that was co-founded by Dr Jerome O’Connell and Prof Nick Holden.
The duo have developed an intellectual property (IP) aimed at drone manufacturers, farm machinery manufacturers, agri-input providers, agricultural enterprises and data aggregators. ProvEye’s technology removes noise from image data, making it easier to understand and derive insights.
Speaking to Siliconrepublic.com, O’Connell said: “Our secret is we can reduce noise, without reducing image quality and sensitivity. That is key.”
ProvEye’s solution
“Many existing solutions fall into the ‘pretty picture syndrome’, where you get what looks like a clean image, but during that cleaning process, much of the information about the condition of crops has been de-sensitised or removed,” the co-founder said.
According to O’Connell, ProvEye can process big data streams across near-field, drone and satellite platforms.
He said: “Our software opens up unprecedented levels of accuracy in the monitoring and assessment of crops, meaning informed and actionable decisions can be made on the management of that crop, using AI-based modelling approaches.”
The company also works with multinationals and SMEs to develop cutting-edge remote sensing and geographic information system (GIS) tools applied to various use cases through its consultancy service.
“Our flexibility is our strength as we assess the needs of each customer on an individual basis, with no one-size-fits-all solution,” O’Connell said. “Much of our consultancy work has been in food or biodiversity monitoring with widespread applications across the world.”
Founding ProvEye
For the last 15 years, O’Connell has been working in remote sensing. He completed a PhD in UCD on the use of remote sensing for mapping Ireland’s peatlands – a project that was funded by the Environmental Protection Agency (EPA), and received validation from NASA along the way.
O’Connell said: “Being from a farming background, I could see the potential of remote sensing in agriculture, especially as machinery options became more digitised and data hungry, and uncrewed aerial vehicles (UAVs) and proximal sensors became more affordable.”
After his PhD, which was supervised by Holden, O’Connell spent four years in the UK studying the use of remote sensing in sustainable agriculture, looking at how UAVs and satellites can be used to measure sustainability at the field and farm.
“It was here that I discovered the shortcomings of UAVs in agriculture, in particular with respect to poor image quality,” O’Connell told Siliconrepublic.com. “It was, and still is, a huge barrier to the widespread use of UAV imagery in industries like precision agriculture.”
He reached out to Holden and began to discuss whether some of the technologies developed as part of his PhD might work to tackle the issue of poor data quality in UAVs. O’Connell said that they then reached out to Enterprise Ireland for commercialisation funding and were successful in late 2016.
Outside of ProvEye, Holden is Professor of Biosystems Engineering in the UCD School of Biosystems and Food Engineering; fellow of the Institute of Soil Science; and president of the International Soil and Tillage Research Organisation.
He began working on optical sensing of soils in 1989 using mainframe computers and migrated to writing his own programmes for the Apple Macintosh when it was launched to begin his work on remote sensing.
O’Connell said: “His early remote sensing work was developing tools for optimising management of industrial peatlands, which progressed to natural peatland change detection, water modelling, landscape degradation and more recently, precision agriculture.”
The journey so far
Since ProvEye was launched, it has gone on to engage with a variety of B2B customers across its different market segments, with a large number of trials completed or currently being completed.
The start-up is launching the next beta version of its software (version 0.4) in the coming weeks, and O’Connell said that it has been optimised for the cloud. Through the consultancy side of the business, he said the firm has built a “much-needed” additional revenue stream.
O’Connell also said that the firm is actively seeking investment and is in advanced talks with an accelerator programme, as well as other sources of investment.
Looking to the future, he said: “The ultimate goal is to be the defacto B2B image processing pipeline across our identified market segments. We have a very clear product development roadmap that will move us from baseline image-processing, right through to decision support and predictive analytics using AI and agent-based modelling for agriculture and other industries.”
He added: “The potential of remote sensing and AI analytics is only beginning to be realised and it could play a large part in our day-to-day lives in years to come, provided it’s implemented correctly.”
Want stories like this and more direct to your inbox? Sign up for Tech Trends, Silicon Republic’s weekly digest of need-to-know tech news.
