Much is being made of how autonomy is going to transform the way in the world operates and agriculture is certainly one industry which will be affected, or so we are told.

Over the last year or so we have seen a good number of launches and announcements from companies, both large and small, all claiming that field work autonomy is just around the corner.

Opening the starting gate

Yet, so far, just one major manufacturer has publicly declared that they are to place an autonomous tractor on the market in the near future, and that is John Deere.

Earlier this month the company claimed the tractor is all ready to go, it’s now just a question of driving it to the field, pointing it in the right direction and flicking the ‘on’ switch.

Yet the question remains as to whether full autonomy is really feasible now or even in the near future. It is notable that the Deere unit itself has to be delivered to the field under manual supervision, vehicle automation for public roads is, it would appear, a long way off yet.

Nexat autonomous tractor — Getting an autonomous vehicle to the field will require a driver for a long time yet

Since the launch of the tractor at a consumer electronics fair in Las Vegas, John Deere has cooled expectations over on this side of the pond with a statement (below) that clarified the situation with regard to European sales:

“John Deere will work on the availability of the technology for further machines and attachments. The launch of the brand in Europe is currently not planned, as safety regulations currently do not allow the use of autonomous vehicles.”

This statement confirms two issues that need to be overcome before it can be sold more extensively: The first is the limited scope of its operations and the second is that EU legislation is not conducive to unmanned vehicles.

Expectations meet reality

Looking back at the predictions given a few years ago there is a something of a gulf in what was projected for the immediate future of machine autonomy generally, and the stage of development and acceptance that we have actually reached today.

Some of the most optimistic forecasts were those of Elon Musk who cheerfully suggested that by 2019 we would be able to whistle up a Tesla from across the country and it would find its way to the door all by itself.

Such a scenario is still in the distant future, despite Elon’s genius at making reusable rockets land all by themselves on a barge out at sea. Rocket science, it seems, is child’s play compared with driving a family saloon down the high street.

Autonomous robotic vehicle — Although launched in 2016 the Starship Technologies delivery robot is still mainly confined to on campus deliveries

Also notable in the sphere of transport is the change in tone from the bullish drive for full automation to the much more gradual adoption of ‘Advanced Driver Assistance Systems’, or ADAS for short.

These are individual components of full autonomy that may be installed in cars without going the whole hog. They represent an admission that we are still a long way from that goal, especially as talk of their full adoption suggests 2042 as a reasonable target.

Tractors as part of a bigger picture

So what is the significance of this delay in implementation for agricultural systems, which are widely touted as being ripe for automation?

To help answer that question we might look to other areas where autonomy is being attempted in a bid to reduce costs and manpower, and one of the most notable areas is defence.

Here in the west we have two basic approaches being made: One of the first out the box was the United States Navy who, 20 years ago, commissioned two classes of smaller inshore vessels which were to be a test bed for reduced manning and, eventually, wholly autonomous operation.

US Navy Zumwalt destroyer — The US navy had high hopes for its lightly manned Zumwalt class destroyer. The project was eventually cancelled

Cost overruns were only to be expected but what was not foreseen was the huge workload placed on the reduced number of crew. Even in 2010, crews were already complaining about getting only six to eight hours sleep for every two days at sea.

Reduction in crew size was also a feature of the US navy’s Zumwalt class of destroyers. It was assumed that halving the number of personnel aboard would greatly reduce costs, yet only three of the projected 32 vessels were built, due in the greater part, to their expense.

Limiting ambition may work

If large scale autonomy has failed to deliver its promise at sea, the UK boasts some success with the second approach of taking baby steps towards fighting wars with less personnel on the front line.

A top RAF officer recently commented that “our drone test squadron, 216 Squadron, has proved beyond doubt the disruptive and innovative utility of swarming drones”.

This latter approach is obviously more modest and with the smaller scale of ambition, there appears to come an increased degree of success.

A valid comparison?

Back in civvy street, these more modest advances are mirrored by the increasing number of video drones and solar-powered robotic lawnmowers which are entering everyday life.

The natural comparison is with large, adapted tractors versus smaller, purpose built field robots. However, the cost of either option is not yet generally known as manufacturers of both are somewhat shy of putting a figure to their products.

What we do see is that the areas where autonomy is starting to work tend to be in a much more disciplined work environments than the open field. Orchards and large-scale fruit and veg being the obvious examples.

It is becoming clear that autonomy is far more difficult than first realized. The advent of GPS and automated steering hailed a bright new future, but that is now getting bogged down in the details of actual implementation.

Servicing autonomy

A further problem arises with the question of who is actually going to service all this electronic equipment once it is out there?

Again, we can look at the history of personal computing and note that getting equipment repaired and software fixed is not always straightforward or cost effective, hence devices have become disposable as they have become cheaper.

This may be acceptable for small robots but the JD 8R 410 costs €430,000 before bolting an autonomous control system to it. That is an investment, not an incidental expense.

Dealers and manufacturers are already sounding alarm bells over the lack of younger people choosing to become technicians, and while there is the rather light hearted belief that a teenager will always be able to fix things it may the case that said teenager prefers not to, unless the money is right.

service technician tractor — Running and servicing tractors remotely is still a distant reality

It is also a generally held belief that the muck and unsociable hours associated with agriculture are a huge bar to recruitment, but is there any industry nowadays that doesn’t expect weekend working from technicians?

The difference may simply boil down to the fact that farming pays less and those that are skilled and motivated will naturally gravitate towards better paid positions.

Stalled by expense

Farmers therefore face the prospect that in a bid to replace the human at the wheel they will be faced not only with a huge capital expense, but also much higher ongoing costs – autonomy is not going to be cheap, certainly not in the short to medium term.

The auto industry has already backed off immediate autonomy and is talking of a much longer period of easing into it gently – it is very likely that agriculture will be doing the same.

Also Read:
Tobroco-Giant to build Doosan’s new loader range

Technicians

The trouble with autonomy in agriculture