Implications of a machine economy

What autonomous machines could mean for sustainable agriculture

If robots can autonomously plant, nurture, and harvest crops, we don’t need humans to operate machines. The implications of this on labour, energy use, and machine design make this form of farming way more efficient. Not just from an economical perspective but from an ecological perspective too.

A short history of automation in agriculture

Automation in agriculture went from hand tools and ox plows to mechanization like Jethro Tull’s seed drill or the grain cradle. Mechanization truly kicked off efficiency. Within a century the the grain cradle alone decimated the amount of labour needed to cut, rake, and bind an acre of wheat1.

Mechanization in agriculture has brought us tremendous wealth. Because we were able to reduce labour from over 55% of the labour force to as low as a few percent in the past centuries, we have been able to securely feed our population, while also having time to invent new stuff. At least in the western world.

With the arrival of automation, the share of the labour force working in agriculture has dropped significantly in Europe. Source.

In developing countries the percentages can still be pre-industrial2. Increased automation in agriculture worldwide would mean an immense boost to prosperity for mankind. That seems like a no-brainer. However using the methods we use in the western world everywhere is a bad idea.

Take the bad with the good

Mechanization has brought many benefits – higher yield, more reliability, lower cost – but it comes with disadvantages too. Besides displacing unskilled farm labour, it can cause environmental degradation. In a previous post we explained the importance of topsoil.

Using heavy plows3 for instance can lead to serious environmental degradation. Plowing can improve soil conditions in the short term, and reduce weed, insects and pests. In the long term however, the tillage results in soil compaction. It also prevents soil organic matter from building up in the soil. This leads to soil erosion. Combined, compaction and erosion make the soil unable to handle rainfall. The effects on the whole ecosystem are devastating. No tillage has its own disadvantages of course. But people are working on solutions, like developing autonomous bots – usually called weed whackers – that help keep weeds in check. This means less herbicide is needed.

What is a weed whacker?

A Weed Whacker is a robot that mechanically whacks the weed at an early stage. This gives crops space to grow, and improves yield. The method does not use herbicides, and does not damage crops. Weed whackers can provide a higher yield, with (even) less manual labour.

Odd.Bot Weed Whacker autonomous robot
The Odd.Bot Weed Whacker provides a higher yield with less manual labour. It contributes to less use of herbicides.

Its elegant manner of working however is just the beginning of the benefits. Weed whackers are much lighter than tractors. Less weight means less soil compaction. Because the bots don’t require constant human supervision, they don’t have to be extremely effective. Time in this case costs less money. This ‘patience’ allows the weed whacker to charge at times the sun shines, which makes a completely sustainable solar or wind powered system possible.

Current developments

Fully autonomous weed whackers are still in development. Prototypes have proven feasibility of the idea, but it will still take some years for a marketable product is ready4. Odd.Bot‘s solution uses a sort of blender to pull out weed root and branch. It then grinds it to bits and the debris will function as green manure5.


This method sounds very elegant, but success depends on the bot correctly recognizing weeds. Humans already have difficulty recognizing weed. For a robot this is even harder. For instance changing lighting conditions hinder object recognition. A solution is to let the weed whacker work at night, under standard light conditions. Another possible solution could be hyperspectral imagery6, which is an automated, non-destructive and rapid alternative to explore plant traits. Hyperspectral camera’s might also be of benefit to recognize insect herbivores.

This example shows the ability to identify individual tree species, but also certain characteristics like stress. Data can help autonomous bots determine what action to take. Source

The video above explains what hyperspectral imagery can show. There’s a hidden world beyond the spectrum that we can see with our eyes, and it could help tremendously with making agriculture more sustainable and resilient.

Connection to digital ecosystems

Solving the problems with the camera alone is not enough, because solid algorithms are essential too. This is where concepts like IOTA’s Industry Market Place are very beneficial. Such decentralized market places open up the data to help discriminate between plants, adding value from compounding data. A helpful feature of this data sharing is that progression of invasive species can easily be tracked, skyrocketing concepts like the Invasive Alien Species in Europe app.

Beyond weed whackers

Weed whackers are hardly the only possibility for autonomous robots to contribute to sustainable agriculture. Other examples are FarmBots. If a machine does not have to carry a human around, and have to utilize this human’s time effeciently, it can be much smaller. It allows for it to take longer to harvest, and could potentially harvest selectively. It would then only pick the ripe fruits, again based on hyperspectral imagery. In the same way, seed planters could be smaller too. They could plant in different patterns, allowing for optimized sowing strategies.

Another possibility is even more exciting: combining light robots with the ability to distinguish various plant species would allow for mixed crops. Monoculture depletes nutrients, leaving soil weak and unable to support healthy growth. Polyculture mixes different crops, imitating the diversity of natural ecosystems. In such a system, weeds would not be pests by default anymore, but instead be part green manure at the very least.


Autonomous machines will be a huge force powering the future of sustainable agriculture. They not only reduce cost, while preserving soil. They also require little or no pesticides, in particular glyphosate. With more and more governments banning it, finding an alternative to pestcontrol is vital to food security.

I like that bots can solve these issues we’re dealing with right now, even though there is still a lot of work to be done. At least it shows we can break the cycle of ever increasing pesticide use and soil degradation, without endangering our food supply.

(The reduction of labor causes problems for people currently working in agriculture. However, I think it is time for us to consider Keynes’ 15 hour workweek, and try to keep ourselves occupied in other ways. More on this in a later post on the future of work.)

About the author

Bas van Sambeek

Communications strategist, specialized in the positioning of high tech products. I help innovative start-ups and fast growing companies in high tech industries clarify their story through positioning and storytelling.

Add comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Implications of a machine economy