R&D Centre - QZ Solutions – Projekty badawczo-rozwojowe

We have the official status of Research and Development Centre granted by the Ministry of Entrepreneurship and Technology

We acquire essential data for conducting research and its analysis thanks to multispectral images and RGB made during flights on fields using the cameras installed on drones.

R&D influence on the company

Our experience in R&D projects

It may seem that in the times of new technologies the innovations are so common that we come across them at every turn. However, the history of many companies in particular sectors, including giants, shows that it is easy to go off the market without thought-through investment in R&D.

As a company dealing with R&D work on a daily basis we present to you the offer of using our services in spectral research, machine learning (ML), artificial intelligence (AI) and making advanced unmanned flights with drones, and also data and image processing.

In QZ Solutions we run a few parallel projects within R&D activities. First of all we conduct research in spectral data analysis for the needs of precision farming.

Our experience and abilities are used in several projects subsidised from The European Regional Development Funds in the total amount of over PLN 45 mln. The most important activities in R&D that we are currently running are using the most recent discoveries from the scientific and information technological world in order to improve crops quality and productivity.

Below we present the detailed description of possible works and examples of projects that we run.

Ground tomato cultivation

In the first place we focus on research in ground tomato cultivation within which we cooperate with one of the biggest food manufacturer in Poland and we create the most modern solutions for planters. Using unmanned aircrafts supplied with spectral cameras we are able to specify the condition of the cultivation basing on the vegetation index such as e.g. NDVI.

We use algorithms to process images in order to estimate the size of crop so the agricultural producer could plan the harvest and sales of the raw product. Thanks to one of the most renowned in the scientific world equipment to make field spectral measurements we identify the early stages of plant disease and enable the farmer to precisely dose plant protection products. Thanks to that the losses caused by such diseases as fire blight, Anthracnosis, Early Blight or Blotches will be lower and tomato cultivation will be more ecological as a result of lowering the quantity of chemical products used.

Cultivation demand for water

Within the project subsidized by The National Research and Development Centre that will start off in January 2020, we will research the possibilities of using hyperspectral imaging and artificial neural network to specify the level of cultivation demand for water. This research will be conducted in cooperation with research institutes specialising in artificial intelligence algorithms, potato physiology and also with practices producing about 50 thousand tons of potatoes each year.

The aim of the project is to create the intelligent system enabling precise potato cultivation irrigation basing on the current soil moisture without using in-site data, such as local moisture sensors.

Sugar Beet Disease detection

We run research in early sugar beet disease detection, first and foremost early stage of beet root cercospora that accounts for even 43% of crop loss.

We conduct the research using unmanned aircraft equipped with hyperspectral camera as well as hyperscpectral spectrometer for ground measurements.

We also run research in the scope of detecting weeds in cultivation with particular regard to beet root cultivation.

Other activities

We also conduct the following within other R&D activities

Spectral analysis of other cultivations in order to detect diseases,
High quality ortophotomaps using unmanned aircrafts, high resolution cameras (20 MPix) multispectral cameras for determining the condition of the cultivation,
We implement algorithms for image detection and artificial intelligence models, first of all for the use of plant disease detection, crop estimation, use of plant protection products and specifying other crop parameters,
Researching technology condition in precision farming.

Technology

Technology that we have:

Multispectral camera integrated with unmanned aircraft

Multispectral camera that calculates vegetation index of cultivated plants. In our company we use this solution to monitor cultivations in proper plants development.

Flights with such cameras are also used to build image sets that are used in the process of algorithm development of artificial intelligence prepared for precision farming needs.

Multispectral imaging allows to collect information about plants vegetation, water distribution and problems with irrigation, as well as agriculture loss and many others.

Hyperspectral camera integrated with unmanned aircraft

Hyperspectral platform is based on 6-rotary unmanned aircraft equipped with 150-band hyperspectral camera. This camera enables image registration in high spectral resolution within the range 470 nm to 900 nm and frequency of 65 frames per second. .

The camera uses spectrometer in order to register changing conditions of lighting. It allows to gain the proper radiometric ortophotomaps even in changing lighting conditions during the flight. Such solution significantly improves effectiveness increasing the number of days in a year in which you can make the flight. The equipment is used for advanced analysis of agriculture cultivation and research related to connecting the hyperspectral vision technology and deep neural networks

RGB camera with high resolution integrated with unmanned aircraft

Unmanned platform is equipped with a camera registering images in high resolution, useful in the process of data collection and ortophotomap building (images with geographical references). It allows for further precise location of registered objects.

We further on use collected images to make analysis, prepare vision algorithms or also in the process of training algorithms of artificial intelligence. Among the applications you can list cultivation analysis in terms of weeding, precise pointing of different objects on fields or detecting agriculture losses.

Hyperspectral spectrophotometer

Spectrophotometer used in our company allows making measurements with a large spectral resolution (1 nm) and in a wide spectral range (400-2500nm), unavailable for other equipment. It is worth noticing that we have mobile versions of the equipment that enables conducting research outside the laboratory – it is used in agronomic, geological and many other researches.

Using spectrometer allows fast measurements and research of phenomenon not visible for other equipment (observed far beyond visible light range). In our research this equipment is used for example for advanced research on early stage detection of disease in tomato cultivation.