How Data Transforms Agriculture

How Data Transforms Agriculture

Every day on my way to work I drive by a lot of fields where farmers are growing plants. Plants for us to eat, for our food to eat and for producing petrol to burn.

The influence of data to make decisions on how you grow plants, which plants you grow and so on will be big in the future. 

Farmers Rely On Gut Feeling and Experience

A typical farm has a lot of acres of land with different soil types and different plants. The farmer needs to know which plants he can plant in which field to get the maximum yield out of the field. That the plant fits the actual soil.

Today this has a lot to do with gut feeling and experience of the farmer.

With this experience he needs to actually plant and grow the seeds. He is going to apply chemicals to support the plants growth.

Applying fertilizer as well as pesticides and fungicides so the plants don’t get sick. When all that is done and the plants are grown he needs to harvest.

He needs machines. The right machines, to harvest everything efficiently and when the harvest is done he needs to sell his harvest. Selling is done on like stock markets.

Every type of harvest has its its own value at a certain point of time.

It all starts with the Internet of Things

To enable insight from data science and big data the first step is to apply sensors in the soil. Then it’s no longer a a case of experience which plants you can plant on which soil.

Knowing the the contents and the quality of the soil from sensors allows you to make smart decisions. Like what to plant when exactly to plant. Or if you need to grow some special plants to fill up the nutrients in the ground, so the next harvest grows better.

There are some companies working on sensors that can do exactly that.

Every field will be connected. Every field will generate data on how is the soil status how are the plants and so on. All this data can be used to optimize then the whole process.

How much fertilizer exactly do you need. How much pesticide and when do I need to apply it. Which agent would be best.

It would be best not to apply any pesticide and so on, but you know, that’s how it works in reality.

Farmers need to optimize. For them, time is money.

They need to be quick, they need to be efficient.

Autonomous Farming

A lot of farmers have huge fields that they are plowing. It takes to actually plough the field, so when you think about autonomous driving this is also a very very large application.

Farmers can just send out the the tractor and he does everything by himself. Or send out the harvester and he harvests. They can concentrate on other stuff.

Harvest Timing

To optimize the harvest you need to find the right time to actually harvest. If you harvest a few days late too late maybe you have some rain it’s too cold and the harvest gets bad.

Or you cannot sell it fast enough and so because the demand is low you geta low price. There are many factors on when to actually harvest and this can and will be optimized by using data from the fields.

Weather data, data from the logistics chain, data from the stock exchange and so on. This will help help farmers a lot.

Optimizing Machinery

The other thing is when they have the data they can optimize the machinery they buy. Buying oversized machinery is very expensive and quite useless.

It also costs a lot of money to maintain. They can use data to optimize what equipment to buy and how to use in a optimal way.

I know this this is all very integrated and influences each other. But I think this is the way to go and that we will see a lot more of that in the future.

Predictive Maintenance

Another thing is maintenance. It’s bad when the harvester is broken.

When the time comes to harvest and that that freakin thing is is not working the farmer is losing a lot of money. In many cases here in Germany the smaller farmers they don’t even have a harvester.

Farmers are renting the harvesters from other farmers. They are dependent on time schedules, so if that thing doesn’t work they have a real problem.

That’s where predictive maintenance can really help. Early detection of problems or prediction of remaining running time before maintenance is key here.

Trade Prize Predicting

In the end farmers need to sell their crop and make money. I know the the farming industry gets money from the government, but nevertheless farmers need to earn money.

That’s why they need to optimize the selling process. When they harvest they need to sell it as quickly as possible before it goes bad.

A friend of mine, who is actually a farmer, told me that they are selling their crops on like a stock market. Each type of crop has a value based on the quality, time and demand (and more).

So today’s price can be a lot better than tomorrow’s price and tomorrow’s price can be worse than day after tomorrow. They are always gambling.

The goal of such a prediction would be to optimise the selling process to make the most amount of money.

There are even insuring companies who offers some kind of a safety net. This way the farmer gets at least a minimum price if everything goes wrong.

Insurance companies will also profit from getting actual data from the whole industry. With that they can limit the the risk of ensuring by predicting the crop quality and price.

Data Platforms and Services

It all starts with with the Internet of Things. When the sensors are cheap and the Internet connection is always there.

Data platforms who can manage the data can deliver a lot of value to farmers. Farmers cannot build these platforms themselves. Hell, even if they could they would not want to do this alone.

The need is definitely there. The pressure is rising for farmers, because selling the crops doesn’t get enough money and they need to optimize everything.

They only have a limited manpower. When you look at it almost nobody wants to do farming anymore. That’s why the automation of everything needs to get higher as well.

More Use Cases?

As you can see there are a lot of applications for using data in farming. It would be very interesting to know.

Do you know someone who is already working on some use cases or a platform for agriculture?

Write me a comment or hit me up on social media.


Say Hello On: Instagram | Facebook | Medium | Twitter | YouTube | Snapchat

Listen to my Podcast on: Anchor.fm | Soundcloud

Subscribe to my newsletter: Here


Damien Sousa de Gouveia

Asset Performance Management Lead for Cohesive EMEA | AI Community of Practice Lead | Enabling IoT Based Monitoring and Prediction Solutions for Smart Asset Management

6 年

Very interesting! I've been discussing something similar with a commercial feedlotter in South Africa. The trick in feedlotting is similar to what has been mentioned. The purchasing price of cattle and the selling price of meat are ,in effect, governed by two distant markets - cattle market and consumer level meat market. Each of these markets ,in South Africa are free markets. The problem is that selling prices and purchase prices are forever shifting and sometimes moving apart. Worse yet, cattle feed has a large maize component to it, which as you have already established, fluctuates in price. However, different breeds of cattle convert food into body mass at different rates in their life time. Think of it as an investment fund that gives you a different interest rate every month. The interest rate starts off slow, increases and then decreases after a while. Where the analogy differs from our discussion is that your variable cost eventually eats away at any profit margin. Where the relevance to your post lies is that using historical data, we can start to predict the rate at which food gain starts to decline, select profit able breeds, attempt to forecast market prices and find peculiar trends - the sky is the limit!

Jacques Bikoundou, MS.

Gen AI & Blockchain Technologist | Data Analyst | Outsystems Developer

6 年

Andreas, I look at the production of crop with a lot of interest. The world of agriculture is a world of time series. There's an element of demand forecast with crop production. Forecast can help those farmers with the expected demand for their crops over time. I am looking at building a model for this.

Benjamin Wielgosz

Making Ecosystems Material to Supply Chains

6 年

Its not all IoT & Precision Ag... that will only allow you to optimise production within a single farm-unit... https://www.dhirubhai.net/pulse/6-agri-business-use-cases-solved-through-mobile-2-silicon-wielgosz/

回复
Gideon Kruseman

WP3 leader One CGIAR initiative on Foresight and Metrics to Accelerate Food, Land and Water Systems Transformation

6 年

https://bigdata.cgiar.org #BigDatainAg is a platform that aims at harnessing the capabilities of big data to enhance the impact of agricultural research.

Julien Heiduk

Senior Data Scientist at ADEO | Expert in Recommendation Systems | Python, GCP, Generative AI, Machine Learning, Deep Learning

6 年
回复

要查看或添加评论,请登录

Andreas Kretz的更多文章

社区洞察

其他会员也浏览了