Can AI save Rioja’s next vintage?

Rioja is using artificial intelligence to turn its fortunes around after a perfect storm of challenges threatened to knock the Spanish wine region off course, writes Lauren Eads. 'On the rocks' is not a phrase anyone wants linked to their wine region – but lately, that’s been the bitter reality for Rioja. An excess supply of wines, a cost of living crisis and a downward trend in global wine consumption have been threatening to knock one of the world’s oldest and most traditional wine regions off-course. Stocks of unsold Rioja have been building in cellars since 2018, leading to unsustainably low grape prices and a drop in profits. Rioja isn’t alone in facing these challenges. Australia, France, California and Chile are all grappling with oversupply. France has spent the equivalent of around US$200 million to destroy excess wine, and has plans to tear up 30,000 hectares of vines to fix the problem, while Australia too is ripping up millions of vines to reverse the glut. Speaking to the drinks business last year, Fernando Ezquerro, president of Rioja’s Regulatory Council, said “nothing is being ruled out”. But hope remains of a rebalancing in Rioja, and much of this surrounds the thorny issue of yields. The DOCa has banned new vineyard plantings for 2024, 2025 and 2026, and also limited production yields to 90% of the official maximum for 2023 and 2024. Last summer, it approved the distillation of 30m litres of surplus wine, and in January this year it introduced a voluntary incentive for winegrowers to reduce their vineyards by 5% as part of its recovery plan, targeting more vigorous red varieties less than 30 years old. But it’s also taking more technological steps to improving Rioja’s fortunes – using artificial intelligence (AI).

Human touch

Typically, yield management has been monitored physically – by someone visiting a vineyard, assessing and making adjustments. This human touch is still important, but since 2023 Rioja’s Control Board has been able to monitor all 66,902ha of Rioja DOCa vines using satellite imagery, and employs AI systems to collect and process the data. The DOCa is using this information to create a predictive model with more than 1,600 parameters, supplemented by historical production data gathered from across the region, including climatic information, vine behaviour and phenological data, in addition to “multispectral” satellite image analysis of each of the 115,000 registered plots.

“Thanks to this detailed analysis, which reveals patterns and trends, we can predict the yields of individual plots with a high degree of accuracy, although this is always accompanied by on-site verification carried out by the experts on the Control Board’s technical team,” explains Pablo Franco, technical director of the DOCa Rioja Control Board. “Being able to cover the entire Rioja vineyard area represents a significant qualitative leap compared to the previous model, which involved partial sampling and randomisation, and could not cover the entire appellation due to limited resources,” he adds. Franco stresses that this tool is aimed at improving quality, rather than addressing oversupply in any significant sense, but it remains an important tool in identifying over-vigorous plots to help restore balance. >

So how does it work?

A production forecast is made around two to three months before harvest. If a vineyard’s yield is looking too high, the DOCa can provide growers with recommendations of specific adjustments, giving updates at key stages of the vines’ vegetative cycle. “This enables us to take proactive measures, which help ensure that production remains within the quality parameters set,” adds Franco. All of the data collected by the DOCa has been available to winegrowers since the 2024 harvest, but can now be accessed via a digital platform as well. “The aim is to provide real-time communication with winegrowers, so they can better plan and organise work in the vineyard,” explains Franco. It’s a powerful tool and a significant undertaking. But how optimistic are producers that AI will improve Rioja’s outlook? Marqués de Cáceres already maps its vineyards using satellite imagery to measure the NDVI vigour index of a vineyard to determine which plots to harvest when. Previously, it would have been necessary to study different maps in order to make a decision, explains Manuel Iribarnegaray, technical director. “AI is an investment in the future, which helps wine-producing areas to evolve and meet consumer needs,” he says. AI has a key role to play in managing the sustainability of vineyards too, believes Guillermo Regorigo, brand manager at Bodegas Ysios in Rioja Alavesa, which is already using AI systems to support decision-making in the winery and vineyards. “Thanks to AI, winemakers will be able to more accurately identify the needs of each plot, thus reducing the use of inputs like water and pesticides, while promoting more sustainable agricultural practices,” he says. “Digitalisation and innovation are key to ensuring the competitiveness of the wine industry. The wine and viticulture sector must have the best tools to help make the best decisions every day and create detailed, real-time vineyard maps.” The difference now is that all Rioja winegrowers, big and small, and those without the financial capability to invest in AI, can benefit from the data shared by the DOCa to improve the management and quality of their vineyards. It’s a democratisation of information, but there is still some caution.

Fears and reluctance

“The winegrowers of our region have always been open to lines of improvement that help both them and the wineries,” says Juan Carlos Ezquerro, vineyard manager at Bodegas Muriel in Rioja Alavesa. “We find hope when we talk about the potential of these tools for the global improvement of the quality standards of the vineyards, but there will always be fears or reluctance, as there is still a lot of ignorance about the subject [AI] by most in the sector.”

Some fear that the information obtained could identify “low-quality” or “high-production areas”, which could lead to future exclusions from the DOCa, adds Jairo Fernández, winemaker at Pagos del Rey in Rioja Alta, which works with thousands of growers. “At the level of individual winegrowers, I understand that initially they will be more reluctant to use this technology. But the winery that buys the grapes from them, or even the regulatory council, will probably use AI and apply it in their vineyards, so these winegrowers will have no choice but to adapt to this new tool if they want to deliver their grapes to DOCa Rioja wineries,” he says. Grupo La Rioja Alta has embraced AI by setting up its own internal department in 2022 – called ‘AI Factory Team’ – which works to implement AI solutions in marketing, accounting, winemaking and the vineyard. But Guillermo de Aranzabal Bittner, deputy general director and sixth-generation family member, acknowledges that it remains in the minority. “I don’t think small producers consider AI to be important for them,” he says. “Most of them don’t have the capital, tools or interest in changing the way they work. Dynamic producers, however, are embracing it. AI will affect all industries – we shouldn’t think ours is any different.”

Hope and caution

J García Carrión, a huge producer that makes wine across 10 Spanish DO regions, is working to implement an AI system in its vineyards, but its director of oenology, Toni González, is sympathetic towards the fears of smaller growers, describing a “mix of hope and caution”. “It’s a promising technology that could bring real benefits, but we need to see how well it can adapt to the complexities of viticulture,” he says. “Things like climate, soil, vines and the unique microclimates of specific areas all play a role. Most importantly, the experience and intuition of the people who work the vineyards is invaluable. One fear is that relying too much on AI could lead to losing human expertise, with machines dictating decisions instead of complementing the know-how of viticulturists. For AI to succeed, it needs to be seen as just another tool in the toolbox, not a replacement.” Using AI to control yields and raise quality is positive, but there also needs to be a healthy market. How do retailers feel about Rioja’s prospects? UK wine retailer Jeroboams has seen a “significant fall” in demand for Rioja wines over the past two years, mostly in the retail side of its business, where Rioja sales have been “gently declining for a decade”, according to wine director Peter Mitchell MW. Its wholesale business has fared better, reporting strong growth in 2022, and only a slight decline this year. “Controlling yields is never a bad thing, and AI could well help improve fruit quality and inform better harvesting decisions, but to re-balance supply and demand, the region needs to reduce its vineyard area (that grew far too rapidly in the 1990s/early 2000s, adding nearly 50% to the area), and try to excite newer consumers with the more exciting wines being made by smaller growers,” he says. “Technology is not a panacea.”

The UK remains Rioja’s biggest export market by a considerable margin, but it’s no secret that many consumers still see Rioja as the reliable, good-value option, rather than something to splurge on. Even at the high end, it tends to be considered sensible rather than extravagant. “Our customers tell us that they feel ‘safe’ with Rioja (our own-label Rioja is a consistent top seller); it’s their reliable ‘go-to’,” says Rebecca Palmer, head of merchant buying at Corney & Barrow. “It seems to be less about excitement than a sense of safety. But I think it depends on the occasion and price segment.” Being regarded as a source of high-quality, ‘safe’ wine is no bad thing, but it doesn’t exactly elevate ideas about the region, or position it at the forefront of people’s minds. Tim Wheeldon, Corney & Barrow’s head of London on-trade, acknowledges the important role that sommeliers play in supporting Rioja’s upward momentum. “Sommeliers are likely to steer customers towards something they perceive as new and interesting,” he says. “Rioja is perhaps considered a safe space, rather than something exciting.” This is despite Rioja making some of its most exceptional wines to date, although efforts to promote more modern, innovative expressions of Rioja have faced resistance. “When we have experimented with more ‘modern’ styles, traditional Rioja drinkers have not generally been in favour, whilst it is a challenge to get non-Rioja drinkers engaged with the region, even for styles that are not heavily oaked,” says Mitchell. Rioja is producing exciting and diverse wines of the highest quality, but if consumers don’t buy into it, preferring to stick with traditional, oaked styles, or gluggable reds, how can it widen its appeal? There’s no easy answer. Rioja will always hold an important place in the UK market, but it has a fight on its hands if it is to not only rebalance supply, but also reignite curiosity among consumers. As for AI, it can’t solve all of Rioja’s problems, but it does address several challenges and, for a region so firmly entrenched in tradition, it’s a bold leap into the future.

Smart tech: How else is AI shaping Rioja?

The impact of artificial intelligence (AI) is also being felt beyond the vineyard. In 2023, Rioja’s Control Board started using AI to support the sensory evaluation of DOCa wines by its tasting panel. “By using AI algorithms to identify patterns in tastings conducted by experts, we can develop more accurate and consistent wine profiles,” says the DOCa’s technical director, Pablo Franco. “This not only supports wineries in their improvement, but also strengthens consumer confidence by ensuring standardised quality.” Could winemaking benefit from greater AI supervision? Jairo Fernández, winemaker at Pagos del Rey Rioja, believes in the collaborative use of technology. “The use of AI in the winery can improve the quality and aroma of the wine, even increase its longevity,” he says. “It will also allow us to make wines defined by AI (specific to the tastes of consumers), making DOCa Rioja wines more competitive in different markets.” But machine algorithms will never fully replace the skill of a winemaker, adds J García Carrión’s Toni González. “Winemaking relies so much on the five senses, especially taste, and AI doesn’t have taste buds (for now),” he says. “It’ll be interesting to see how the technology develops and how it might support winemakers, without losing the human touch.”