Contact Us

Information

[email protected]
Contact Us

Month: June 2023

Illicit plants control platform

Posted on by admin

Description

Detection of illicit cannabis plants with the help of spectral cameras mounted on drones to create maps of illegal cannabis plantation.

Problem

Illegal plant cultivation remains a big problem in society. Such plants can be grown in the backyard or in the middle of a field. Existing methods for detecting such plants are far from ideal. They may be accurate, but only work in close proximity to the plant, or they may be able to work over a large area, but with extremely low accuracy.

Solution

Plants emit different light spectra depending on their metabolic processes, which allows scientists to distinguish them as a separate group. The use of spectral cameras has become a popular tool in the mapping of the surface of an area under investigation. These cameras are able to capture and analyze the light emissions from plants and other objects in the area, allowing for the creation of detailed maps that reveal the distribution and types of vegetation, as well as other features of the area. 

However, the data obtained from these maps is often complex and difficult to interpret. To address this issue, researchers have developed neural networks that are specifically trained to process the spectral data collected by the cameras. By using these networks to analyze the data, it is possible to obtain more accurate and detailed information on the condition of the surveyed area, such as the health and growth rate of the plants, and the presence of invasive or illegal species. These techniques have proven to be invaluable for researchers and conservationists who are working to better understand and protect the natural world.

Acoustic diagnostics of insect pests

Posted on by admin

Description

Acoustic diagnosis of insect pests is based on devices capable of detecting vibrations associated with the vital activity of insects inside plants.

Problem

Insect pest infestation can lead to colossal damage to crops and trees. Current methods of detection can’t provide status of infestation irl and must be checked by the human being, looking for the signs of damage. Furthermore, when combating said insects, people don’t localize insecticide spread, because of poor data collection. Thus leading to unneeded ecological damage.

Solution

Borer insects produce vibration upon digging through the structure of the plant. Those vibrations are caught by acoustic devices mounted on the plant stem. Given data sent to the server, which collects data from the web of devices. Using AI, data is summarized to be more manageable to the end-user. Data given to the user provides all information on location and severity of infestation in real time. Thus leads to less intensive insecticide spreading, lesser ecological damage to animals, water and crops and fewer consumption of insecticide.

Edible cricket farming

Posted on by admin

Description

Alternative food source for humans based on the use of crickets. These insects have high nutritional value and are relatively inexpensive to breed.

Problem

The population is steadily growing, and so is the demand for food products. People primarily get their protein from meat. In addition to the fact that livestock produce methane, which is harmful to the environment and is sometimes released in catastrophic volumes, raising livestock requires huge pastures and fields and a significant amount of water. The cost of maintaining such pastures can be painfully high.

Solution

Cricket is an amazing source of animal protein and many groups of vitamins. Breeding crickets can occur in more confined conditions than traditional livestock. Insects consume very few resources and have a minimal negative impact on the environment. The end product of the cricket life cycle can be protein powder, rich in calcium, iron, and fiber.

Seismic exploration data processing and interpretation system

Posted on by admin

Description

Program that uses artificial intelligence algorithms to build a depth-velocity model of a deposit. The model can be used to analyze and understand the characteristics of the deposit, such as its composition, structure, and porosity.

 

Problem

Deep processing has rightfully earned its place in seismic data processing. Without accurate positioning of layers laterally and in depth, successful exploration and therefore hydrocarbon extraction would be impossible. Against the backdrop of the growing interest in hard-to-recover oil and gas reserves, the use of modern environment reconstruction methods is essential in the profitability assessment stage of drilling.

 

Solution

Program is based on artificial intelligence algorithms, specializes in building a deep-speed model of a deposit. The system finds wide application in seismic exploration, namely for the interpretation of seismic data. Artificial intelligence trained using deep learning is used for this purpose, which allows finding the optimal solution of the initial task independently using non-linear algorithms. Thanks to this program, you will be able to obtain a more detailed and accurate model of the deposit, which will help increase its mining efficiency and reduce exploration cost

Respiratory Infections Control Platforms

Posted on by admin

Description

Promising non-invasive method for diagnosing respiratory diseases. Using a device equipped with a microphone, information is processed to detect anomalies, based on which the program can output a diagnosis.

 

Problem

Coughing is an innate protective reflex that is part of the body’s immune system. It is a common symptom of respiratory problems. Determining the causes of coughing is the most frequent subject of a doctor’s diagnostic search, as it is not always possible to establish the nature of the cough, which affects the results of the examination and further treatment. A crucial issue of global significance is the control of the epidemiological situation and the prevention of the spread of viral infections.

 

Solution

Using technology to determine the probability of a person being infected with a respiratory virus allows for timely identification of potential outbreaks of a dangerous disease. Recording devices allow for continuous sound recording of the environment. In real-time mode, the recording is analyzed for the presence of coughing using neural networks. When coughing is detected, the recording is segmented and the segment is sent for detailed analysis to detect the disease. Based on the prediction result, the system generates a warning about the high risk of a coughing person for those around them.

Monitoring and predictive detection of natural disasters. Earthquake. Landslide. Glacier melting

Posted on by admin

Description

Processing data from satellite images and various sensors via AI in real-time. Forecasting several kinds of natural disasters and informing the population about its affected area, degree of danger, time of effect. Those disasters are as follows: earthquake, landslide and glacier melting.

Problem

A natural disaster can cause loss of life or damage property, and typically leaves economic damage in its wake. The severity of the damage depends on the affected population’s resilience and on the infrastructure available. It is important for individuals and communities to be aware of the risks associated with disasters, and be able to take steps to prepare for these events. Forecasting of hazard events must be the main concern of specialists.

Solution

Collecting data in a surveillance zone using satellite images and, if possible, ground sensors is a critical step in predicting natural and what not disasters. However, processing the huge volumes of obtained data is a challenging task. Real-time use of artificial intelligence allows for more accurate signature processing and quick identification of potential disasters. In addition, 

It is important to alert the population to potential disasters so that people can take measures to ensure their safety. The alert should contain information about the time of the event, magnitude, and estimated start time of the disaster. Thus, data collection, signature processing, and population alerting are important components in preventing catastrophes and saving lives.

National e-voting system

Posted on by admin

Description

Remote electronic voting system protected by cryptography based on blockchain technologies, for people with disabilities. Provides transparent, comfortable and trustworthy experience.

Problem

Classical methods of voting do not inspire confidence among the population and provides little to no ease for people with disabilities. You can only vote at polling stations. For some, this is neither comfortable nor doable. All the votes are trying to be as transparent as possible, but this is not enough, people still do not trust the results. At any stage, a number of vulnerabilities are known: from bribing voters to stuffing ballots when counting votes. 

Solution

Remote electronic voting system with cryptographic protection based on blockchain technologies. In order to vote, an identification system must be set. Votes can be submitted from anywhere, with only limitation – internet connection – thus being accessible to anybody. The solution is built on reliable and proven technologies: blockchain and smart contracts inherently do not allow the possibility of fraud.