- Статьи
- Internet and technology
- Easy work: An AI assistant will help even a beginner assemble a drone.
Easy work: An AI assistant will help even a beginner assemble a drone.
Russian students have developed an AI assistant for the production of drones. The system allows you to assemble UAVs even for users without experience working with drones: the application provides step-by-step instructions, visual cues and elements of augmented reality. According to the developers, the technology helps to reduce the number of assembly errors by up to 70% and can be used as an educational tool for training specialists in the field of unmanned aircraft. The principles of the system's operation are described in the Izvestia article.
Assistant for drone assembly
Russia has created a new generation mobile application with an AI assistant for the production of unmanned aircraft systems. The project was implemented within the framework of the federal program "Personnel for UAS", operated by University 2035.
The application uses augmented reality elements: the user needs to point the smartphone camera at a QR code or a part marker, after which the system shows information about the element, from size to compatibility with other nodes. The built-in TensorFlow Lite neural network allows you to recognize details by shape and work without an Internet connection.
— One of the features of the solution is an interactive 3D model with step-by-step animation of component installation. In addition, the application offers several build scenarios, allowing the user to choose the appropriate option right in the process," explained Andrey Rastopshin, technical leader of the DronAssist development team.
According to him, the system is designed, among other things, for people with minimal experience working with drones. Moreover, visual cues can reduce the number of errors by up to 70% and simplify learning basic assembly operations. In the future, the development will provide the country with new personnel for ALS and simplify the threshold for entry into the profession.
Currently, the developers have prepared a prototype application and a server part with a database. According to the team's estimates, the cost of the solution does not exceed 100 thousand rubles, which is significantly lower than the price of stationary industrial analogues. The application works on a regular smartphone and does not require specialized equipment. According to Andrey Rostopshin, this will make the production of UAVs in Russia more affordable, simple and efficient.
The prospects of technology
The main advantage of the development can be a combination of relatively low cost, mobility and ease of implementation, says Nikita Danilov, CEO of Fly Drone, market expert at NTI Aeronet.
— If we talk about the stated cost, it really looks lower than that of stationary complexes with similar functionality. This can be an attractive option for drone manufacturers, especially for small businesses and training sites. In addition, using such an application potentially allows you to involve less qualified specialists or newcomers in the field of unmanned aircraft in the assembly process. This means reducing the burden on the personnel market and saving on the payroll," the expert noted.
According to him, the advantages of the solution also include the ability to work in the field and flexibility when used with various types of UAVs. At the same time, Danilov drew attention to the possible limitations of the technology.
— There are risks too. For example, problems may arise when recognizing parts, especially if the marking is of poor quality or the components themselves differ from those originally included in the system. In addition, if the application is running on a smartphone, there are limitations related to the performance of a particular device. Therefore, much will depend on the quality of implementation and practical experience of application," he believes.
There is still a high demand on the market for both specialists in the field of UAS and educational tools for their training, said Andrey Korigodsky, CEO of Over.
— Printed instructions or even instructional videos do not always allow you to quickly master the assembly process. This makes learning more difficult and time-consuming. Therefore, solutions using AI and augmented reality can become a useful auxiliary tool," he explained.
According to the expert, such technologies can significantly reduce the time of basic training.
— If, according to classical materials, a specialist conventionally masters the assembly of the device within a week, then using an interactive AI assistant can potentially shorten this process to one day. This is especially true for schoolchildren, students, and aspiring professionals," said Andrey Korigodsky.
At the same time, he stressed that such applications should not be regarded as a full-fledged replacement for comprehensive engineering training.
— This is a promising solution, but only one of the elements of a large training system for the development and production of UAS. Without practice, an understanding of electronics, equipment configuration, and an engineering base, such tools alone will not be able to train a full—fledged specialist," the expert added.
Pavel Kamnev, market expert at NTI Aeronet, believes that the educational sphere remains the most realistic scenario for the application of the development. He considers the use of such applications to be particularly promising in the framework of existing projects of the Ministry of Education and Science and other government initiatives aimed at training personnel for the unmanned aircraft systems industry.
— If we talk about large UAS manufacturers, the prospects here look ambiguous. Drone designs, components, and software are constantly changing, so the application will have to be regularly adapted to new platforms and configurations. This is quite a laborious process," he said.
IT expert Sergey Pomortsev is skeptical about the prospects for mass implementation of such solutions in industrial assembly.
— In practice, the production of UAVs is not only the sequential installation of parts according to instructions. The process raises issues of electronics configuration, calibration, component compatibility, and quality control. It is still difficult to fully complete such tasks with a mobile application. Therefore, such assistants are more suitable for training and initial acquaintance with the industry than for serious production," he believes.
At the same time, the expert does not exclude that in the future the development will be able to occupy its niche in the market if the authors manage to integrate it into a single infrastructure with software and hardware complexes for training in piloting drones.
Переведено сервисом «Яндекс Переводчик»
