Pro česky mluvící odkážu na podrobnější popis v angličtině ale také na skvělý CZPODCAST, nebo MEETUP

CZPODCAST Soundcloud

kde jsme o tom co děláme mluvili. Pak stačí kouknout na pár videí níže a snad je to jasnější.

What problem are you targeting?

Ullmanna is targeting one of the costliest and most burning problems of today’s agriculture: weed control. Recent global estimation of yield loss to weed competition in the fields was 9%, which adds up to annual economic loss of 24 billion euro. Moreover, left uncontrolled, weeds can lead to a 100% yield loss.

The traditional ways of controlling weed - herbicides - have no direct organic substitutes (like e.g. insecticides). Moreover, they are increasingly getting restricted by new policies and their alternatives are becoming costly. Despite all these facts, the use of herbicides keeps stagnating. It will not be possible to meet the commitment of a 50% decrease in their usage presented in the Farm2Fork strategy unless scalable alternatives emerge. In organic farming, on the other hand, manual labour has proven to be both hard to access and scale post-pandemic.

Ullmanna is providing a future-proof alternative to agrichemicals and manual labour.

How does the world work now in relation to this problem?

Currently, the largest market share for weed control in conventional farming is taken by the usage of herbicides via chemical spraying. Thanks to the pressure for sustainable agriculture and consumer preferences for organic food, the market share and relevancy of organic and sustainable farming is growing. In organic farming the weed control is based on a combination of mechanical inter row weeding and manual labour which can cost up to 50% from farmers’ revenue.

There are different approaches to sustainable weed control present on the market. Chemical spot spraying (John Deer, Ecorobotix) is the fastest but simply not environmentally-friendly, as it still employs agrichemicals. Laser weeding (Weedbot, Carbon Robotics) as a promising approach struggles with speed and throughput and entails high adoption cost. Both approaches share the same disadvantage: they kill only visible weeds.

On the other hand, mechanical weeding is using knives moving underground and suppressing the conditions even for yet invisible weeds. The most advanced products present on the market (KULT, Lemken, F.Paulsen and Garford) operate based on colour recognition, which noticeably decreases their efficiency, operation range, and user experience.

How do you solve the problem?

Mechanical weeding on its own is a well-known approach. The focus of its innovation centres around improvement of precision, autonomous operations, and knives control. Series of breakthroughs in colour recognition and robotics have led to several new technological solutions for weed control.

Ullmanna’s invention - AROW - is an AI-driven crop recognition unit, which operates a system for mechanical weeding for one row of crops. AROW is a highly adjustable device which can be used on its own as an integral part of NEWMAN (Ullmanna’ proprietary weeding machine) or integrated in any other commercial solution for mechanical weeding. Typically, weeding machines are equipped with 6-15 AROW units.

With AROW, Ullmanna introduces the next generation of improvement by utilising AI-based image recognition which can recognize crops independently of colour, so the product works even in extremely weedy conditions (green on green) and is capable of processing fields with same crops of different colours (e.g. salads). Moreover, AROW unit doesn't rely on weather or light conditions, and there is no need for the adjustment of parameters during the process. Thanks to Ullmanna's engineering, the weeding machine can carry e.g. 15 AROW units, each covering one row of crops. This makes Ullmanna's product scalable for sustainable and potentially conventional farming as opposed to competition, which often has lower number of units per machine due to their larger size and the need for shades (to negate light effects).

Ullmanna's improvement in recognition processes and operations algorithms also allows AROW units to go toe-to-toe with competition in inter-row weeding in terms of speed, while already being 2.5 times faster in intra-row weeding (AROW was proven to be 100% efficient at 5 km/h compared to 2 km/h competitors’ solutions).

It is important to mention that the knives steering system of Ullmanna is first of its kind and protected not only with patents, but with a private system of collection and processing of the training data for neural networks, which is a significant barrier for the competition.

Finally, the user interface of the AROW unit provides an opportunity to select the crop and then adjust the performance by changing just one parameter ('window' of the knives opening) instead of keeping track of the field conditions, weather, light, etc. (in competitive solutions).

AROW was already proven to be effective and precise. It achieves 99+% precision in recognition and steering for a number of crops (most advanced are sugar beet, chicory, lettuce). Even the current version of Ullmanna’s machine is a feasible alternative to a manual labour of 60 people, as it is capable of processing more than 10 hectares in a single day. In sugar beet, for example, the 12-row setup performs with a speed of 2ha per hour. The potential output could be 40ha per day.

Přihlaste se k newsletteru a posuňte se vpřed!

Přihlaste se k newsletteru a posuňte se vpřed!