“eyeSpot” – leaf specific herbicide applicator for weed control in field

 

Responding to concerns about loss of herbicides and pressure to target pesticides better and in lower doses, research will build on the expertise at Reading, Precision Farm Robotics and Knight Farm Machinery to develop a herbicide ejector which will apply metered droplets to leaves of unwanted plants. The active ingredient must be systemic and should be non-selective (e.g., glyphosate). For commercial application, the herbicide would be formulated to minimize spatter and run-off so that, in principle, herbicide is applied to weeds; not to the crop or soil.

 

The ejector will “point and shoot” high velocity droplets – not a spray and not gravity-fed droplets. Glasshouse research will quantify the dose of droplets required to kill different weeds at different growth stages in variable environments. In the field, using an appropriate platform (e.g. a tractor-trailed system or an autonomous vehicle), field testing will assess actual herbicide inputs, weed control costs and herbicide residues, testing the hypothesis that weeds can be controlled leaf-specifically without crop damage.

 

Key benefits of the system once commercialised could include:

 

Crop	UK area, ha	Herbicide input 2011 Tonnes active (UK)	Adoption of robot by 2022, market share (% of UK area)	Predicted. reduction in herbicide applied, Tonnes UK
Onions	9365	40.21	15	5.72
Carrots	11135	51.26	15	7.30
Brassicas	39643	33.99	10	2.51

Sources: FERA 2011, 2012 and Jayne Dyas, British Growers (personal comm.)

 

• Concerns about loss of herbicide actives and circumventing disincentives for investing in discovery of or obtaining approval for new herbicides are addressed, at the same time providing an alternative strategy for EAMUs. Chemicals for which approvals have been withdrawn may be able to be approved if dose rates can be reduced by one or even two orders of magnitude and if spray drift can be eliminated.
• Longer-term, imagery and data for weed management could potentially be adapted to address other crop management issues such as detection and analysis of crop stress, estimations of growth rate and early yield estimates. Where applicable, maps of black-grass patches may be used for patch spraying of pre- (and post-) emergence herbicides in the following crop with projected savings of £31/ha (cf. Murdoch et al., 2010).
• EyeSpot is an alternative to both selective herbicides and herbicide tolerant (e.g. GM) crops. Even were companies willing to develop new selective actives for vegetables or GM varieties, "eyeSpot” is arguably superior because the system is not beholden to any one herbicide, the only constraints being the two obvious ones, namely that any herbicide used must first show efficacy at dose applied and secondly be sufficiently slow-acting and systemic so that it can be translocated within the plant from the leaf/leaves to which it is applied.
• While use of an autonomous vehicle is not a prerequisite for the system, a further benefit is that it offers UK field vegetable growers the opportunity to get ahead of competitors by capitalising on over 10 years development costs by consortium members in developing the basic robotic platform. Like most autonomous robotic platforms for agriculture described in the literature, the first four generations of AgBot were research machines and, as such, inappropriate for commercial use. This project will utilise (for field testing) a fifth generation pre-commercial platform with additional robustness and safety features being added based on field testing during the project. The basic platform, which will be tested on fields during the project, will be provided as an in-kind contribution by Precision Farming Robotics Ltd. (PFR).

 

Assuming a successful outcome to the research on the “eyeSpot” herbicide droplet ejector in this project, the following route to market is proposed leading to financial benefits to companies involved as well as to Reading University. Precision Farming Robotics (PFR) will license the technology to Knight Farm Machinery (KFM) for manufacture at their factory near Oakham for both the U.K. and potentially for other worldwide markets. Weed identification and control protocol software will be licensed by Reading University but all software maintenance and support and post-commercialisation upgrades would be done by PFR. Overall, marketing, distribution and service in the UK could be carried out by Knight Farm Machinery. Within the first five years of sales, (which given a successful outcome to the project, could commence in 2018) adoption percentages are tabulated above and are estimated to require 319 systems by 2022 to achieve season-long control of weeds in field vegetables and other potential row crops.

Users would purchase the system (estimated cost per unit is c. £25000), one unit being recommended for each 65 ha. Depending on the crop, this area assumes each field would need to be treated up to four times at 2-3 week intervals to ensure control of late germinating weeds. Training would be required but were an autonomous machine the platform of choice, labour inputs and energy usage are expected to be very much lower than for mechanical weed control (e.g. using plant specific methods like the Robocrop) and/or spraying herbicides (post-emergence). Benefits to growers are much greater than just herbicide savings.

With respect to chemical approvals for droplet application, the glyphosate-based herbicides ‘Roundup Energy’ and “Roundup Flex” (Monsanto UK) have recently been approved for an Extension of Authorisation for Minor Use (EAMUs 0354/2013 and 2528/2013) to allow UK growers of various tuber, root and bulb crops to use these products as spot-sprayed, inter-row herbicides.

At the end of the project, much more extensive field testing will be needed to ensure robustness and reliability in a wider range of conditions, soils and cropping systems. 

 

 

1. Herbicide Actives: lack of or potential loss of herbicide actives due to legislation or development of herbicide resistance in weeds; 
2. Environmental issues and opportunities: 

• to reduce off-target spraying;
• to help meet demand for more sustainable crop production with lower pesticide inputs;
• to reduce energy inputs and soil moisture loss by reducing/eliminating need for mechanical weed control both inter- and intra-row;
• to provide an alternative to GM herbicide tolerant crops.

 

1. To develop an automated system (eyeSpot) for projection of metered herbicide droplets to individual leaves of unwanted plants in row crops;
2. To adapt Reading University’s current image analysis algorithms or develop new ones to distinguish weeds from selected field vegetables;
3. To develop algorithms to determine which individual leaves of a weed would be ‘safe’ targets for the ‘point and shoot’ herbicide droplet treatment;
4. To model dose-response relationships of herbicide droplet application system for leaf-specific weed control for principal weeds of field vegetables; and
5. To field-test efficacy of real-time weed control by eyeSpot technology.