Society is undergoing what many refer to as the Fourth Industrial Revolution. Also known as Industry 4.0, this era is marked by the integration of physical and digital worlds, often blurring lines between the two. It includes advancements in fields such as biotechnology, nanotechnology, the Internet of Things (IoT), robotics, artificial intelligence (AI), and quantum computing; AI is at the forefront. What once was the stuff movies are made of is becoming a reality.
What is Artificial Intelligence?
Artificial intelligence is the ability for machines to learn. The human brain learns via sensory information transmitted to our short-term memory through a network of neurons. Once processed in our short-term memory, the brain transfers some of this information to our long-term memory, the vast set of archives where our life experiences are stored. Whenever we acquire new knowledge, it is stored in our memory where we can retrieve it when needed, applying past experiences to current situations.
Similarly, AI learns from experience by using advanced algorithms to identify patterns or features contained within large amounts of data. The primary advantage is AI can process significantly larger amounts of data in shorter amounts of time and with less errors than humans. AlphaZero, an AI designed by research company DeepMind, mastered chess in just four hours after being programmed with the rules of the game, beating world-chess champion program StockFish in a breakthrough moment. Impressive!
There are many theories, technologies, and sub-fields involved with AI, such as machine learning, neural networks, deep learning, cognitive computing, natural language processing, and more.
Companies have used these technologies for years. We find it in government, health care, manufacturing, retail, and sports. It’s not surprising to see the wine industry also taking advantage of these developments. AI and machine learning can be implemented at nearly every stage of the winemaking process — from monitoring vineyards to increasing production efficiency to predicting the quality of wines.
Winegrowing and Production
Wine Australia, has partnered with DigitalGlobe and Consilium Technology to develop GAIA (Geospatial Artificial Intelligence for Agriculture), an AI software that provides valuable insight into the quantity and condition of vineyards across the country.
Using machine intelligence along with DigitalGlobe’s satellite image library, GAIA aims to pinpoint and plot every vineyard in Australia. The high-resolution satellite imagery is able to provide direct information about Plant Cell Density (PCD) and the Normalized Difference Vegetation Index (NDVI). Deep learning neural networks enable GAIA to use this information to continuously improve its classifications of vineyards, crop condition, and more. It can monitor plant health, assess soil moisture, and even determine fruit quality.
Ailytic, a South Australian tech company, has also developed an AI software to streamline wine production using a technique called prescriptive analytics, an area of business analytics designed to find the best course of action for a given situation. The AI collects data from production operations and uses it to analyze variables such as temperature and inventory. Then it generates an optimal production schedule based on the analysis, allowing companies to maximize their time and money.
Predicting Wine Quality
It may seem counterintuitive that a machine can determine whether a wine tastes good. However, the quality of wine depends on a combination of chemical properties that influence taste, flavor, and aroma. Some may consider winemaking an art, but it’s actually quite scientific.
Data science is a field that involves gaining insight and understanding from data. Machine learning is a sub-field of artificial intelligence based on the concept that computer systems can learn from data, identify patterns, create models, and make predictions.
Datasets, such as the one here, can be created based on a wine’s physicochemical composition. Then, using data science, we can extract insights to help us understand what makes a wine taste good. Likewise, a computer system modeled on machine learning can use data science to learn what makes wine taste good and thus predict its quality based on the distribution of its components.
Buying Wine
Some companies have used AI to change the way consumers buy wine. ‘Virtual sommeliers’ like Wine Ring make recommendations according to individual consumer preferences. Simply rate wines by telling the app whether it’s a Love, Like, SoSo, or Dislike. Wine Ring’s algorithm keeps track of your ratings. The more you participate, the more it learns your palate. Snap a photo of a label next time you’re buying wine and the app will tell you if you’ll like it based on your profile.
Hello Vino is another app that uses artificial intelligence to learn your wine preferences. Vivino also uses label recognition technology, but it does not offer suggestions based on preference. Rather it provides instant access to a database of wine reviews and prices. Wine Searcher uses AI to classify wines and link users to half a million products at over 80,000 retailers, a task that otherwise would be incredibly difficult and time-consuming.
The aforementioned are just a few examples of a growing trend. In the past, consumers would consult a sommelier or other wine expert. Today, however, they are increasingly turning to their smartphones. According to a 2015 Survey of American Wine Consumer Preferences , 25% of wine drinkers use wine apps to help them decide which wine to buy.
The Future
The thought may cause some wine purists to cringe, but fully automated vineyards and wineries don’t seem far-fetched. I imagine Matrix-like sentinels hovering around as they monitor and guard vineyards, destroying pests and harvesting grapes with their tentacles; RoboSomms in Five-Star restaurants greeting customers by name based on facial recognition from their visit two years ago. After an evening of drinking, they hop into an AI-powered driverless vehicle to get them home safely… Then again, maybe it’s just my imagination.
Time will tell what the future holds. A popular phrase states “the future is now.” What we visualized yesterday has materialized today. We may have a way to go before RoboSomms and vineyard sentinels, but technology advances so rapidly that this article was outdated before it was published. Who knows what’s in store?
The foregoing article was originally posted on my wine blog. Some of my most popular articles have been reposted to Medium. Interested in more wine content? Click here.