The Future is Fruitful with the Help of Machine Learning
World hunger is on top of everyone’s mind–especially with tomorrow being World Hunger Day. The thought of everyone having enough to eat is a comforting one; One that can soon become a reality. As with the introduction of most new things, there’s some hesitation about how much of a role AI should play in something so delicate as having control over the world’s food supply. Common questions include: what are the potential applications for AI in this industry, will AI automation take away jobs, and is the technology developing too quickly? Despite the hesitation, so many advances are being made towards AI saving our crops, so more food makes it to more tables.
AI can positively impact the entire growing process, from the early stages of identifying weeds and finding new locations to plant crops, to identifying plants dying of disease. Even in harvesting, AI helps by identifying which crops can be used for human consumption and where in the food chain they belong.
Current State of World Hunger
According to Forbes, roughly 820 million people don’t get enough to eat each day, and 1.9 billion are experiencing food insecurity. The world population is greater than 7.9 billion people, and a quarter of the world’s population works in the agriculture industry. So why aren’t we able to grow enough food? 20-40% of crop losses are caused by pathogens, animals, and weeds, and that doesn’t account for the fully grown food that gets thrown out because it doesn’t look perfect, is stored poorly, or isn’t purchased by the sell by date. Nearly half of all edible food is thrown out each year, which is about 2 billion tons.
The United Nations is estimating that in order to make sure everyone has enough to eat by the year 2050, the world will have to increase food production by 70%. A great way to start is reducing the volumes of perfectly good food discarded each year, and AI can certainly solve that problem.
Repurposing Perfectly Good Food
Sometimes food may have a blemish or two, but that doesn’t mean it’s not safe to eat. Unfortunately, large amounts of “flawed” food are considered unfit for grocery stores to sell. Companies like Imperfect Foods sell these foods that have a blemish or two that grocery stores refuse to sell. They’ve managed to save 145,823,731 lbs of food since launching in 2015. They partnered with Topos, a company that leveraged machine learning models to predict which areas would see the highest numbers of customers wanting to buy imperfect produce. But what about foods that have more than just a blemish that people might not want to purchase? A company called Tomra has a solution to this.
Tomra has created sensor-based machines that carefully look over each piece of food and determine if any part of it is usable for human consumption. They leveraged AI to help train their food sorting machines to think like a human would when looking at the food. Their logic is that these foods can be used as an ingredient to another recipe where appearance isn’t a factor. It’s the machines job to determine where each piece of food belongs. This has caused the company to be able to save 5-10% of food per yield, which is the equivalent to 25,000 trucks filled with potatoes.
Plant Identification Apps – Early Warning Detection for Diseases
In an earlier article, AI is the Key to Convenience, we talked about a smarter way to garden through the use of plant identification apps. These apps allow a user to take a photo of any plant and find out what type it is and if it’s poisonous or has a disease.
Farmers are now utilizing this same technology to take photos of their crops to quickly detect if a plant has the disease they suspect or if its ok. By detecting diseased crops in early stages, farmers prevent other healthy crops from contracting it, thus decreasing the amount of food lost each year.
It’s not just diseased crops that farmers need to watch out for. Weeds are also an issue as they suck up valuable nutrients that crops need to grow. Plant identification apps can detect weeds from newly growing crops, so that they can be eliminated before any major damage is done.
Increasing Farming Potential
Crossbreeding among foods isn’t a new concept. Scientists have been doing this for years to create foods that are different colors or remove an undesirable characteristic. The science behind this, however, is evolving. Instead of creating new foods, they are perfecting what’s currently out there.
Carnegie Mellon launched FarmView, an AI system that helps identify each plant’s characteristics. This allows farmers to select only plants that will produce more crops or have a higher disease resistance.
Appen’s Commitment to a Brighter Food-filled Future
The idea of ending world hunger seems more and more a possibility with each passing day, and we’ll certainly need the help of AI to make it happen. Our team and Crowd are currently working on several image classification projects that fuel the artificial intelligence used to power these initiatives. We collect, annotate, and classify data to train AI models on topics including food, plants, transportation, and more. Having this wealth of information allows identification apps to know the difference between a weed versus crop and a healthy plant versus diseased.
We’ve shared the different applications for AI to help through food repurposing, plant identification apps, and the ability to increase farming protentional. These applications will only serve to aid those currently working; Jobs won’t be taken away. Data for the AI lifecycle ensures that the technology is thoroughly tested before being implemented into the food industry.
We also understand that there may be some additional hesitancy of allowing AI to play a role in something as delicate as food and encourage you to reach out to us, so that we can help ease your concerns and share why we think responsible AI can help save the day.
Appen is committed to make a positive impact for all aspects of our environment, and that includes contributing to efforts to reduce or eliminate world hunger. We will be hosting Ted Chats and trivia days on environmental topics for our employees sharing ways to get involved, as well as fundraising money for EarthDay.org. Next week on the Appen blog we’ll be talking more about how AI can help the environment.