Automatisk præcisionsstyring til mekanisk ukrudtsbekæmpelse i gartnerier og planteskoler

Automated Precision Steering in Mechanical Weed Control in Nurseries

Summary

The need for controlling weeds in outdoor vegetables and nurseries is often great because it is a question of competitively weak crops grown at large row spacing, typically 25-50 cm. The growing pressure on the use of pesticides in agriculture on the part of society and the fact that several herbicides have been removed from the market due to reassessment have led to mechanical weed control gradually gaining a footing. In market gardens and nurseries, traditional inter-row hoeing for controlling weeds in the intervals between the rows is primarily being used.

For a long time, agriculture has wished to have the possibility of automatically steering inter-row cultivators in row crops because this can make mechanical inter-row weed control considerably more rational in several respects. Firstly, because an extra wage cost may be saved in the cases where an extra person is needed for accurate steering of the implement itself besides the tractor driver. Secondly, because the working capacity is expected to increase considerably since the driving speed can be increased as probably also the working width of the implement. Thirdly, because the working environment of the tractor drivers can be improved as compared with the existing steering systems, which require great power of concentration and which therefore are limiting to the daily operating time. Automatic precision steering can contribute to a reduction in the consumption of herbicides by combining inter-row hoeing with band spraying.

The purpose of the project has been to study the perspectives and the possibilities of steering inter-row cultivators automatically with reasonable precision and reliability in connection with mechanical weed control in row crops in market gardens and nurseries. Also, to elucidate the possibilities of using extra equipment fitted to an inter-row cultivator for mechanical control of weeds in the actual crop rows and to quantify the effects of controlling weeds and possible damages to the crops. Furthermore, to set out the relevant applications in outdoor vegetables and nurseries, including assessing the consequences for the consumption of herbicides, the consumption of energy, and the grower’s finances.

From the beginning three potential systems were to be compared with manual steering:

ECO-DAN’s LPS (Local Positioning System) Danish vision system (ED)
Frank Poulsen Aps.’s AUTOPILOT Danish vision system (FP)
Robert Tucker AEGIS (Agricultural Guided Implement Systems) Australian system based on aluminium row markers and a metal detector (RT)

The Australian system had to be dropped after some initial tests and technical problems. The other systems were tested in trials with transplanted cabbages, sown cabbages, and seeded onion at Research Centre Flakkebjerg and in pricking-out and seed beds of Caucasian fir, common spruce, and bird cherry at Hedeselskabet’s Nursery, Brønlundgaard. The tests were carried out at different stages in the development of the crops and at different speeds and distances from the rows. The project has been characterised by the fact that – contrary to expectation - the steering systems that we were testing were not operating perfectly and that they were continuously developed during the progress of the project. It has therefore been necessary to improvise during the project as regards the original plans. On the other hand, we have been able to follow and in part contribute to the development of the systems, which has provided us with a good insight in the strengths and weaknesses of the systems.

We may conclude that both Danish systems work just as precisely as manual steering under ideal conditions. That is: when there are symmetrical crop plants in visually distinct rows and weeds in limited size and numbers. Both systems have not been reliable, however, when the ideal situation was not present. At the beginning there were problems with light-shade contrasts in the scanning fields, but both manufacturers have solved this during the first year of the project. The major problem has probably been distinguishing between weeds and crop, especially if the weeds have been of the same size as the crop plants. The experiences with both steering systems show that dicotyledonous crop plants need to have a leaf area of 2-5 cm² and to stand at a regular distance in the row for the camera to be able to recognise the row. In arboreal crops and onion, the plants must be higher than 5 cm for the systems to be able to recognise the row and steer the inter-row cultivator just as well as manual steering. Generally, it may also be concluded that the smaller the crop plants, the more sensitive the systems are to irregularities in the rows and to weeds between the rows. If weeds sprout and grow just as fast as the crop plants, supplementary preceding control methods that do not depend on having to distinguish visually between crop and weeds become necessary. A third problem has been asymmetrical crop plants, e.g. lopsided plants (especially at transplanting), overturned plants or plants sown in broad bands. When the steering systems react to such plants, the implement may damage plants in the other rows. Despite these problems there is a perspective in automatic steering systems because it has been elucidated what the systems can and cannot so that they may be incorporated into integrated control strategies. The systems are still being developed, and it is estimated that in a few years they may replace manual steering in a large number of situations in agricultural crops, vegetables, and arboreal crops.

The report contains an analyses of the economic advantages and disadvantages that automatic guidance systems involves as compared to standard manual steering systems. First of all, an automatic system needs to be as precise and reliable as existing manual systems. Automatic systems that cause severe steering errors will easily turn economic advantages into significant and unacceptable losses. Particularly in nurseries, losses can be very high even at a rather small level of damage. However, an assessment of the current status of damages in nurseries caused by steering errors with existing manual systems revealed that sometimes up to 20% of the plants were significantly damaged and thus useless for sale. Generally, the economic benefits of purchasing automatic steering systems in the advantage of manual systems increases with the number of passes that requires an extra person for steering that automatic steering is actually replacing. Also increases in the working capacity through higher driving speeds and/or the use of wider implements will improve profitability. The release of working time that automatic steering may bring makes it possible to add more labour to other productive tasks on the farm. Generally, the profitability of purchasing new equipment for automatic steering increases with the annual area where the system is going to be used. Purchase costs are particular important for the overall weed control costs at small areas to be treated, whereas areas of more than 200 ha annually makes purchase costs insignificant. It has not been clarified to which extent automatic steering can replace laborious hand weeding of intra-row weeds, but in case the economic potential for using automatic steering becomes very high.

Automatic steering systems are estimated to make significant reductions in herbicide consumption more possible in most vegetable crops. Notably in conventional cabbage cropping where mechanical weed control is widely used and based on hoeing technology that requires accurate and rational steering. In other vegetable crops, such as direct-sown onion and leek as well as carrots, reliable automatic steering will make band-spraying much more practicable than current steering systems. Band spraying is expected to reduce herbicide consumption by at least 70-80% in vegetables grown in single rows at 50 cm row spacing. Automatic steering is only expected to have minor impact on herbicide consumption in nurseries because most of the usages are soil-applied herbicides applied early in spring. However, automatic steering is estimated to make hoeing operations closer to the plant rows more likely, whereby time consumption for hand weeding might be halved.

Automatic steering can lower fuel consumption for mechanical weed control operations, if the working rate can be increased. In addition, automatic steering is expected improve current thermal weed control methods by targeting heat application only to the areas, where it is intended to be effective. By achieving that, energy consumption can be lowered by 60% of the usage known for current application methods, such as broad flaming and steaming.