Agricultural robotics and drones: diversity of functions, forms

Our research, published in Agricultural Robots and Drones 2018-2038: Technologies, Markets and Players, describes how robotics is transforming the world of agriculture. Indeed, our previous articles have described agricultural robotics and drones could become a $28Bn and $35B industry by 2028 and 2038, respectively. In this article, we seek to demonstrate the diversity of functions, forms and fortunes that exists, and will continue to exist, in agricultural robotics.

 

To this end, we will discuss all manners of robots and drones including basic and intelligent tractor-pulled robotic implements, automated guided vehicles, autonomous tractors (level 3 to 5), fleets of unmanned agrobots, automatic harvesters, stationary as well as mobile robotic arms, mapping as well as spraying drones, and so on.

 

The images below demonstrate this diversity. Here, you can see that tractors are becoming increasingly autonomous; that fleets of small unmanned agrobots are being developed; that implements are being intelligent essentially acting as tractor-pulled computers-on-a-wheel; that drones are moving beyond simple mapping to conduct additional functions such as spraying; that strawberry harvesting robots are evolving in their form factor; and that both stationary robots and AGVs are used in dairy farming.

Traditional robotics is mostly about stationary robotic arms performing a repetitive task in a known and constrained environment. Agriculture and dairy farming are also no different. Indeed, one of the primary use case of robotics in this field is in robotic milking. Here, specially-designed rugged robotic arms locale the target destination by projecting a pattern of known dots and measuring the degree of distortion upon incidence. The robotic arms then manoeuvre the suction head, which are essentially the end effectors, to do the milking. This is already a demonstrated and successful significant market (>1$Bn) in this field.

 

 

The use of robotic arms is also evolving in agriculture. Here, the trend is also towards mobile picking. The ultimate application here is in fresh fruit harvesting in environments like orchards. This task is still largely manual and even the tools that aide the human pickers have barely evolved over the past fifty years. The slow pace of progress towards automation here is testament to the complexity of the task at hand: the robot must identify and localize the fruit in a complex and varying canopy structure, it must the plan its path and manoeuvre the robotic arm, and it finally must rapidly yet gently pick the fresh fruit.

 

Despite the complexity, innovation is notable here. Companies are developing novel end-effectors based on a suction head, a rugged low-cost cutting tool, or a combination thereof. The picking algorithms are also evolving. In the past, specific instructions had to be scripted whereas now the trend is train the robots using deep learning on a dataset that can either be synthetically generated or be obtained through large-scale trial and error operations. In parallel to these, companies are seeking to utilize soft robotics to enable the end effectors to mechanically adopt the shape of the fruit. This is a novice field and there is still much room for improvement in terms of better control of the pressure distribution (logic system for fluidics).

 

When it comes to fresh fruit harvesting not all are waiting on the ultimate prize. Others are tackling lower hanging fruits. For example, we are already in the second generation of strawberry harvesting robots whilst some are others are seeking to autonomous mobile robots that transport the picked fruit thus enabling the dexterous human to focus exclusively on picking. In our report, Agricultural Robots and Drones 2018-2038: Technologies, Markets and Players, we provide technology and market analysis/forecasts for the use of robotics in milking as well as fresh fruit harvesting. We take both a short as well as a long view in our projections and technology roadmaps.

As with the rest of the robotic world, agricultural robots and drones are also becoming mobile and increasingly intelligent. Let us first consider the increase in intelligence. A good example is to be found in tractor-pulled implements. Today implements predominately only perform a mechanical function and only very few, often in organic farming, are equipped with basic vision technology. The basic vision technology here offers simple row following capability to help provide automatic position adjustment.

 

The technology is now evolving. The newer generations are essentially multiple ruggedized computers and camera systems integrated into the tractor-pulled implement. These systems take images as they are pulled along the field and identify weeds vs. crops. The systems then control a precise de-weeding mechanism, e.g., precisions spraying of the right selective herbicide, to take site-specific actions.

 

 

The vision technology here is much more powerful than the simple row-following approach. Here, deep learning algorithms are trained to identify crop vs. weed and to later differentiate between various weed types. This is no easy feat and cannot be achieved towards using traditional script-based approach towards programming. Simply put, this is because one is dealing with complex and variable entities that change shape and appearance during their growth.

 

The technology here has already been demonstrated and indeed commercial activity is accelerating. For example, last year we witnessed the first major start-up acquisition with a price tag exceeding three hundred million dollars. In our report, Agricultural Robots and Drones 2018-2038: Technologies, Markets and Players, we provide technology and market analysis/forecasts for the use of robotics, including basic as well as intelligent tractor-pulled implements, in weeding, harvesting (e.g., lettuce thinning) and other tasks. We take both a short as well as a long view in our projections and technology roadmaps

Now let us consider autonomous mobility. First, we have automated guided vehicles (AGVs). There are vehicles that follow a guidance infrastructure such as magnetic tape or regularly placed reflectors or bar codes. There are commonly used in factories and warehouses, essentially acting as a distributed conveyer belt. They are also used in this sector. In dairy farming, for example, heavy (e.g., 500Kg) automated vehicles are used in barns to push the feedstock closer to the animals or specialised cleaning AGVs are employed to clean the manure.

 

 

The navigational technology in agriculture has also evolved past rigid infrastructure-dependent automation. Indeed, as we have written previously, agriculture is the leading adopter of autonomous mobility technology. Tractors with level 3 and 4 of autonomy are common sights in many territories. The technology itself is transitioning towards master-slave and fully autonomous versions. The primary function is of course to boost output whilst cutting headcount.

 

In parallel, the advent of unmanned mobility can give rise to agricultural vehicles with new form factors. In particular, large fleets of small, light, slow agrobots may come to replace a few large, heavy and fast traditional agriculture vehicles. These new vehicle form factors are possible because autonomous mobility eliminates the wage bill overhead per vehicle. In this case, a few persons will remotely monitor and control the operation of a large fleet. Here, the productivity of individual units may be lower than traditional powerful vehicles such as tractors but the overall productivity- at the level of the fleet- could be higher.

 

In our report, Agricultural Robots and Drones 2018-2038: Technologies, Markets and Players, we provide technology and market analysis/forecasts for AGVs in dairy farming, autonomous tractors (level 3 to 5) and unmanned agricultural robots (data scouts, weeding and multi-platform robots). We outline the latest developments and provide short-, medium- and long-term (2038) forecasts in unit numbers and value (with or without the cost of tractor itself).

Consumer drones have come a long way. The market for drones (hardware platform) has in fact become commoditized and the business scene has been largely consolidated with DJI and its aggressive pricing emerging as the current victors.

 

 

This business dynamic has forced many to seek opportunities in the enterprise sector. Here, the focus is on prosumer or professional drones and/or software-based services. Agriculture is still a primary target market even though the players are now learning the peculiar challenges of this sector.

 

On the hardware side, the specific needs of agriculture have left room for rugged low-weight fixed wing designs. The drones have also started to become larger, seeking to carry out functions such as precision aerial spraying. Here, gasoline-powered unmanned helicopters can cover a larger field but come at a higher price, whilst larger (professional) multicopter battery-powered drones do the opposite.

 

On the software/service side, the first use case is in aerial mapping. Unlike satellite or plane-based aerial imaging, drones can provide close-range and regular data at low costs, employing this tool to even smaller farmers. Here, drones equipped with simple multi-spectral cameras can yield many useful indices about the health and growth status of different.

 

The interest here has now shifted towards the provision of services and data analytics. Companies are now seeking to offer more than the commonplace NDVI map, focusing on the provision of specialized indices or actionable advice. As with rest of the drone industry, as hardware become commoditized, the value in agricultural will shift towards subscription-based services and analytics and opportunities will open up for niche specialized lightweight sensors. In our report, Agricultural Robots and Drones 2018-2038: Technologies, Markets and Players, we provide technology and market analysis/forecasts for drones. Here, we provide short-, medium- and long-term (2038) forecasts in unit numbers and value for mapping and multi-functional (e.g., spraying) agricultural drones, unmanned spraying helicopters and drone-based data services.