Increased sustainability in Norwegian cereal production using inter-row hoeing by means of high precision auto steering

A tractor with auto-steering is available, and equipment for implement steering is mounted. Further, a tool for testing the accuracy of the steering is constructed, consisting of a 3 m wide bar, with three tines, each having a depth-regulating wheel, attached. Reference lines in form of a steel wire was placed in the field by means of a laser. Comprehensive testing of steering accuracy has been performed. We tested auto-steering alone and in combination with a tractor independent implement steering. For each system, we tested three different corrections for the GNSS-system: Differential GNSS (D-GNSS), RTK-GNSS with CPOS-correction signal obtained from the Norwegian Mapping Authorities (via the GSM-network), and RTK-GNSS with correction signal from a local base-station. All combinations were tested at three different speeds, at 3, 7 and 10 km/h. The conclusion was that RTK-GNSS gave higher accuracy than D-GNSS, but the difference was negligible between the two correction systems for the RTK-GNSS. This was probably due to a favourable distribution of the signal stations of the Norwegian Mapping Authorities in this region, and/or sub-optimum location of the local base-station during testing. The auto-steering of the tractor appeared to be so good (standard deviation: <2.9 cm) that additional implement steering improved the results only marginally (standard deviation: <2.3 cm), particularly at the lowest speed. Auto-steering of the tractor alone gave a standard deviation of 2.8 cm at 3 km/h, whereas the corresponding deviation for auto-steering plus implement steering was 2.3 cm. Regarding the camera-based steering system, there has been a change in the project plan. After project start, the British company Garford Ltd. has entered the marked with a camera-based system designed to handle 12.5 cm row spacing in cereals. We thus found it most expedient to use resources originally planned for the own system development to gain access to the Garford system, in order to maximize the progress of testing camera-based solutions for auto-hoeing in cereals under Norwegian conditions during within the project frame. The equipment was tested during the cropping season 2015, where we were primarily exploring effects of forms of the hoeing blade/tine (three types were tested) and driving speed (3, 6 and 9 km/h). Additionally, we developed a system (a wheel-carrier with three cameras) for quantifying soil translocation resulting from of the hoeing-process. The conclusion in 2015 was that the steering of the hoe works with sufficient accuracy, but that stones in the soil pose a challenge, which have to be solved for the equipment to work on morainic soils. During the winter 2015/2016 we therefore developed a stone-release-system designed for the Garford hoe. The sales manager of Garford was shown the equipment in April, and was very impressed. The equipment was tested in 2016, both in field trials and in normal fields. The results showed that one pass with the hoe reduced the number of weed plants by approx. 42%. As in 2015, the weed pressure was generally low with very little perennial weeds. This contributed to a non-significant effect of treatments on yields. In June, the hoe was tested on the farm of Gjermund Ruud Skjeseth in Ilseng focusing on hoeing in side-slope terrain. The hoe was tested in five runs with increasing sideways slope (4, 6, 7, 8 and 9%, respectively). The conclusion was that the hoe handled side-slope poorly. The first excursion in the project was arranged in May 2014. The trip went to the research farm of the Copenhagen University and neighbouring fields, where demonstrations of numerous auto-hoeing devices used in cereals with double row spacing were performed. In August 2014, the project was presented and auto-steering demonstrated for 130 visitors at Apelsvoll during the annual Field Day. On a similar Field Day in 2015, the steering accuracy of the Garford was demonstrated with great success. In 2016, we presented the hoe with it?s new stone-release-system for 110 visitors on the annual Field Day at Apelsvoll. In December 2014, we arranged a successful seminar, with external speakers, both from Norway and abroad, in addition to internal contributions. The concept was repeated in April 2016, with more than 20 satisfied participants. During the winter 2016/2017, the project and the most important results have been presented at various occasions, such as the thorough article in ?Norsk Landbruk? (REF) and the speech at the Kornkonferansen 2017.

There is a need for non-chemical and integrated weed management strategies in the cereal production. A promising method here is hoeing. Inter-row hoeing can control larger weeds, the risk of crop damage is lower and the method is less sensitive to treatme nt timing than the more common weed harrowing, using a spring tine harrow. So far, existing, inter-row hoeing systems applicable to cereals rely on a wide row spacing, which reduces yields. In this 3-year project we intend to design a prototype steering s ystem for inter-row hoeing in cereals, suitable for normal (12.5 cm) row spacing. Three systems will be developed and evaluated; 1) automatic steering according to the position of the tractor along a virtual line, 2) enhanced automatic steering using an a dditional GPS receiver on the hoeing implement, and 3) automatic hoe guidance using camera systems. Different combinations of geometrical shapes of the hoeing blade and driving speed will also be tested in order to maximize performance in terms of weed re duction without damaging the cereal plants. The best hoeing steering system combined with the best hoeing blade x speed interaction will be tested under realistic conditions. This represents the first step towards a marked ready package (steering, adapted hoeing blades and hoeing platform), which may contribute to reduced herbicide usage in conventional cereal cropping, and represents a realistic, non-chemical alternative for the farmer in cases where resistant weeds are a challenge, and for organic farme rs. The current project is part of a German-Norwegian cooperation. In an independent, but similar, parallel project, the German partner (Hohenheim University) will focus on winter cereals, while the Norwegian partners will focus on spring cereals. The par tners in the Norwegian project, which will be led by Bioforsk, are Felleskjøpet Agri SA, Adigo AS and the Norwegian Agricultural Extension Service. Stakeholders will be represented in an advisory board.