Professional Engineering Magazine : Green-fingered robots

Green-fingered robots

Fleets of small robots could one day be seen out in the fields planting seeds and harvesting crops. Heath Reidy reports

We are nearing that time of year when a charity song with the words “Feed the world…” is likely to be heard in most supermarkets. But agricultural experts say that current issues, such as the world’s growing population, along with the recession and climate change, are making it increasingly difficult to do what Band Aid’s hit demands.

Farming giant: Claas 600 Lexion combine harvester can change speed and direction with no input from the driver – but it would be too risky to put a machine of this size in the hands of a robot alone

Agricultural engineer Professor Richard Godwin says there are fears that the recession could affect food supplies and see prices rise. He says it is also becoming difficult to find people to do agricultural jobs, with countries such as China seeing more people leaving behind rural life to find better-paid jobs in cities.

One of the solutions, Godwin says, is to look to engineers to find quicker, smarter and cheaper ways of growing and harvesting crops. It makes sense to use the machines that are playing a part in most industries today – robots.

Robots are proving to be a cost-effective way to aid food production, from slicing up vegetables for salads to packing products for distribution. So is there room for robots in agriculture too?
Simon Blackmore, agricultural engineer and managing director of business UniBots, says there is. “We have got to find more efficient ways of supporting farmers and allowing them to become more efficient by using automation, rather than using people,” he says.
Blackmore says that using robots in agriculture will save money and improve production, as they can operate without a driver and work day and night.

But let’s forget new technology for a moment. How far are we from robotising today’s agricultural machinery? Well, a long way – decades, in fact. The 600 Lexion combine harvester, for instance, with a 12,000-litre grain tank, is claimed to be the largest machine of its type. The combine, which uses a GPS system, was developed by global firm Claas. It is used to harvest a range of crops, and it can change speed and direction without any input from the operator. But, despite how close it may seem to becoming driverless, the combine couldn’t be adapted to work on its own, as a robot.

Claas product manager Laurence Rooke says that the firm could not risk leaving a machine of its size and power to roam free around a field. “It is not feasible to do it,” he says. The risk of an accident if the computer that controls the machine malfunctions is too great.

And the firm is not alone in its opinion. Godwin says that, while combines are getting very sophisticated, the probability of them becoming driverless in the next 10 years is “remarkably slim”.

While he says it is not “technically difficult” for engineers to solve problems such as robots hitting obstacles in a field, by using sensors for instance, it would be a question of “being prepared to bite the bullet”.

Blackmore hopes to resolve the issues of using robots in agriculture from the developments within his firm UniBots – a conduit for agricultural robot designs and prototypes from universities around the world. Past developments include small robots that can cut grass and pull out weeds, as an eco-friendly alternative to pesticides. But now the company is looking to engineer a family of small robots that will play an even bigger role.

“We think we can now develop a new set of agricultural machinery based on small smart machines,” says Blackmore. “At the moment all agricultural machinery is getting bigger all the time. We are trying to reduce the scale of treatment right down to individual plants.”


Growth area:Artist's impression of the UniBot machines carrying out different tasks

Working from a GPS system, the 2m-long wheeled robots, which will be made from automotive parts, will perform more complex farming tasks. These include planting seeds, by using a set of rotating tines on a vertical axis fitted underneath the robot. The machine can stir the soil to make a small hole and drop a seed into it, before another tine covers up the hole.

A second robot will do what is called crop scouting. Using laser scanners and ultrasound technology, the robot will record information and images of plant quality, so the farmer can map out areas of a field to see where pests and weeds are growing.

Selective harvesting will be a larger robot’s main task, whereby the farmer can select certain crops for harvesting, rather than a whole field. Blackmore says: “We would be able to sort the products in the field before we harvest them, rather than after harvesting them.”

The 3mph robots’ three main factors are that they are small, slow and smart, which “addresses the safety issues as well as we can,” says Blackmore. But he says the main challenge will be getting the machines to work alone for long periods. His team is designing “all of the appropriate behaviours” to enable the robots to do this.

A related development is the robotic shepherd, a device that a person can use to oversee as many as 10 robots at a time. Although it means that someone is still required to work with the robots, it would resolve another of Godwin’s issues – maintenance – as information about the machines will constantly be relayed back to the robotic shepherd, so that if there is a problem it can be spotted.

Blackmore says: “We would have continuous updates on what the robot is doing and how well it is behaving. When one gets stuck or has a flat tyre, a person is close by to try and help sort it out.”

Powering the robots so that they are energy efficient and can run for a lengthy duration will also be a challenge. Blackmore says that batteries do not contain enough energy for the robots to work for a long time. The machines are therefore likely to have hybrid engines, using a combination of electricity and diesel.

Blackmore is also looking into using biofuels. In the future, he hopes the robots will run on fuel cells.

The size and speed of the robots may to some extent resolve concerns, such as over safety, but Godwin says it may create problems in production and flexibility, compared with the capabilities of larger machinery. “The bigger the machine you can have, the more productive it can be,” he says. “The question is, if you let loose a fleet of small robots in a field, in 24 hours, could they have the same productivity as one big machine?”

Even if the robots overcome these issues, it may be difficult to find a market for them. Blackmore says that none of UniBots’ work is targeted at Britain because there is “no interest” in it. He says that most of his research is aimed at the rest of Europe and America.

But, as global issues of climate change and the recession grow, Blackmore says that the British government should take the idea of using robots in agriculture seriously. “We should be having discussions about self-sufficiency,” he says. “It is better to eat local, fresh organically grown produce, and the robots are ideal to help us do that.”

The use of agricultural robots is even more unlikely in undeveloped countries. As well as it being a problem to find skilled workers to maintain the machines, such countries will find it difficult to find money to buy the technology.

Godwin says: “Although I think that these technologies are not going to be that expensive, they are still going to be outside the reach of the absolute peasant farmer.”

Using robots in undeveloped countries could also displace jobs, which residents rely on to sustain their economy.
That said, there are countries that are interested in using robotics in agriculture.

Some of Blackmore’s robots have been used in parts of Europe, such as Greece, and he says that people in China showed great interest in using the technology when he visited the country to discuss its potential last month.

If other countries follow suit, there may well be a strong future for UniBots’ technology in agriculture. But the same cannot be said for big robotic machines, unless more research is done.

“What we need to do is find development money to see how far we can go with it because the advantages are clear,” says Blackmore.

© PE Publishing, 26 November 2008

www.profeng.com The home of Professional Engineering on the web

Professional Engineering

Related links and resources

Green-fingered robots

www.profeng.com The home of Professional Engineering on the web

Professional Engineering

Search our back issues

Related links and resources

Green-fingered robots