Machine learning and artificial intelligence can provide the answer to one of the world’s most pressing questions: how to tackle world hunger.
Thomas Malthus will be forever associated with the topic of world hunger. His 1798 Essay on the Principle of Population predicted a grim future where populations would continue to grow exponentially while food supplies will be restricted by finite land.
His prediction may have been incorrect, but the concerns about world hunger are still relevant today. Despite the huge strides in technological advancement the question still remains, how do we stop people from starving? The United Nations estimates that we will need a 70% increase in food production to feed the world by 2050.
And the problem is not solely with the production of food. For instance, there are concerns over access to water, the distribution of food around the globe and the impact of conflicts which could affect supplies.
Fortunately, thanks to innovations in machine learning and artificial intelligence (AI), we are closer to finding an answer to world hunger than ever before.
Tomatoes for your salad or for ketchup?
A good place to start when it comes to feeding the world is to maximize the current output. According to the UK’s Institution of Mechanical Engineers, as much as half the world’s food (or two billion tonnes) ends up in the trash.
TOMRA Sorting Solutions uses AI to detect the smallest amount of food in order to reduce waste. Essentially, the technology aims to view food the way we do; instead of merely separating food into “good” (A1 Grade for supermarket sale) or “bad” (any abnormality in size color or appearance) which are often discarded, the aim is to optimize each and every crop. Potatoes not considered worthy of gourmet plates could still be used to make french fries or chips, or perhaps donated to food banks.
The TOMRA machine uses Near Infra-Red spectroscopy to analyze the molecular structure of products while technologies such as x-rays can measure its elemental composition.
Using these techniques, the technology recovered 5-10% of produce – or 25,000 trucks of potatoes each year.
Tackling plant disease
Another source of inefficiency in food production is the health of the crops. Approximately 20-40% of crop losses are caused by pathogens, animals and weeds. AI can help farmers quickly detect diseases and work out the best course of action.
Biologist David Hughes and epidemiologist Marcel Salathé are using deep learning to analyze 14 crops which are susceptible to 26 diseases. When a computer was fed 50,000 enhanced images of crops it was able to correctly diagnose 99.35% of them. With regular photos, the figure falls to around 40% but that number will climb.
Once it improves, it will be used to power the app PlantVillage which allows farmers around the world to upload pictures of their plants for diagnoses.
Roomba for weeds
It is not just the health of the crops but the quality as well which is crucial. In the US for example, farmers use 310 million pounds of herbicide on corn, soy and cotton fields. As it is eloquently described here, rather than a minimal approach to pesticides it is more along the lines of carpet bombing.
One solution to this is the LettuceBot which uses machine learning and a database of more than a million images to photograph 5,000 young plants a minute and identify each sprout as a lettuce or a weed. Once found, it sprays a strong fertilizer to kill the weed and fertilize the lettuce in one go.
LettuceBot’s company Blue River says its machine can reduce farmers’ use of chemicals by 90% and already supplies 10% of the lettuce in the US each year.
World hunger: ending the famine crisis
Africa’s famine crisis is the gravest emergency since the Second World War according to the United Nations with six million people at risk of starvation in Somalia and 14 million more in South Sudan, Nigeria and Yemen. In addition, one in four people in sub-Saharan Africa is malnourished.
The Nutrition Early Warning System (NEWS) is one solution which is hoping to solve the root causes of starvation. It uses big data and machine learning to detect early signs of food shortages such as crop failure, droughts and rising food prices. This information is used as an early warning to help tackle impending crises.
There has already been some success with 170 farmers in Colombia avoiding potentially crippling losses after machine learning algorithms revealed a drought on the horizon. Consequently, the solution advised farmers to skip the planting season, saving them more than $3m in input costs.
Rather than judging the impact of artificial intelligence and machine learning on how well the likes of Siri respond to our voice commands, it is through tackling some of the world’s most pressing concerns that the new technologies will make a lasting impact.
Inbenta utilizes its patented natural language processing, artificial intelligence customer service, and +11 years of research & development to create interactive chatbots with an industry leading +90% self-service rate.
Companies around the world including Pinterest and Docusign utilize Inbenta to maintain a personal service for their customers while reducing support tickets.
Interested in finding out more? Our team of experts is at your service to design a custom proposal for you.
Check out our similar articles
- What Is Predictive Search? Reasons & Tips to Implement It
- Chatbot UX: 5 Ways to Keep Customers Engaged
- 10 Cool Chatbot Examples to Inspire Your Project