fbpx

MowHawk camera detects invasive plant species and litter

13-01-2021 | |
Photo: MowHawk
Photo: MowHawk

Dutch company Datacadabra developed a smart camera system called MowHawk. The system will be used for verge management and is able to recognise and collect data about invasive plant species. It also recognises litter. In the future, MoHawk could be used in agriculture, for recognising nests or weeds in grasslands.

The new MowHawk (or: mowing hawk), is a smart camera system that is mounted on the cutter bar and recognises plants or litter. The MowHawk was developed for the Dutch province of Overijssel and will be mounted on two verge mowers this year.

Detect invasive species

The MowHawk is able to detect invasive species such as the giant hogweed or the Japanese knotweed. These exotic plants do not traditionally occur in the Netherlands and are harmful to people and nature, says Renee Roeleveld, business developer at Datacadabra. Once the location of these plants has been registered, a municipality can take targeted action.

Text continues underneath image

The MowHawk was developed for the Dutch province of Overijssel. In 2021 the camera will be mounted on two roadside mowers. It's to provide more biodiversity, circular use of the grass clippings and insight during the entire verge mowing process. - Photo: Datacadabra

The MowHawk was developed for the Dutch province of Overijssel. In 2021 the camera will be mounted on two roadside mowers. It’s to provide more biodiversity, circular use of the grass clippings and insight during the entire verge mowing process. – Photo: Datacadabra

MowHawk is self-learning

Observed images are quickly compared with images in the system’s software, and an algorithm determines whether such a camera image corresponds to an unwanted plant. According to Roeleveld, this makes the camera self-learning.

Besides weeds, the camera also recognises litter. The MowHawk camera sends the images to a roadside manager. The manager can use the data to determine exactly where exotic species grow and where there is a lot of litter.

Clean roadside clippings (free of litter and invasive plants) can be processed and disposed of regionally as a soil improver in agriculture or as an addition to a fermenter.

Text continues underneath video

MowHawk introduction from Datacadabra BV on Vimeo.

Recognise nests and weeds in grasslands

The solution has been developed for innovative roadside management, but the MowHawk can also be used in agriculture. It is technically possible to teach the camera to recognise nests of meadow birds and weeds in grasslands.

If the camera registers something out of the ordinary, the location is registered using GPS. This means that if the MowHawk recognises a nest, the farmer has to take action himself.

Map weeds

Another practical solution is to use the MowHawk to map out where there are a lot of weeds in grassland. The working width of the camera is adjusted to the width of the cutter bar, so the highest possible resolution is available.

Text continues underneath image

The camera system detects litter, invasive plant species and the weight of the clippings. As a result, the composition of the grass clippings can be classified, processed to a high standard and used in a circular manner. - Photo: Datacadabra

The camera system detects litter, invasive plant species and the weight of the clippings. As a result, the composition of the grass clippings can be classified, processed to a high standard and used in a circular manner. – Photo: Datacadabra

Price

In addition to the MowHawk camera, Datacadabra also supplies the software for processing the data. The camera system consists of four modules for roadside management (Vision, Tracking, Base, Cloud) which can be purchased for a total sum of € 25,000. This includes annual service, maintenance and calibration (€ 7,500) and license fees (€ 7,500).

When the MowHawk is adjusted for agriculture, the price will have to be re-determined. This depends on the module that is needed for agricultural use.

The observed image is quickly compared with images in the system's software, and an algorithm determines whether such a camera image corresponds to an unwanted plant. - Photo: Datacadabra

The observed image is quickly compared with images in the system’s software, and an algorithm determines whether such a camera image corresponds to an unwanted plant. – Photo: Datacadabra

Looman
Robin Looman Machinery writer for Trekker