Logistics is a key element of the successful work of modern companies. With globalization, the rapid growth of e-commerce, and increasing supply volumes, effective supply chain management is becoming extremely important. Artificial Intelligence (AI) offers innovative solutions to optimize logistics processes, automate routine tasks, and accelerate delivery. However, with the capabilities of AI, new challenges also arise. In this article, we will look at how artificial intelligence is changing logistics and what problems may arise during its implementation.

Artificial intelligence is actively used in logistics to solve a number of key tasks, such as route planning, warehouse operations management, and demand forecasting. Let’s look at the main areas where AI demonstrates its effectiveness.

One of the most important aspects of logistics is the organization of transport routes. The more efficiently routes are planned, the faster and cheaper the delivery of goods. AI is able to analyze large amounts of data on road conditions, traffic congestion, weather conditions, and customer preferences to select optimal delivery routes. AI systems can determine the shortest routes in real time, taking into account traffic jams, redirect vehicles when the traffic situation changes, or estimate delivery times and update forecasts for customers[1].

Companies such as Amazon and UPS are already using AI to route cargo, which can significantly reduce fuel costs and increase customer satisfaction.

Warehouse operations is another area where artificial intelligence is showing excellent results. AI technologies make it possible to automate inventory management processes, track the movement of goods in a warehouse, and predict the need for inventory replenishment. Smart warehouses use:

• Robots for moving goods: I-guided robots move goods around the warehouse, reducing the need for manual labor(Fig.1).
• Automation of sorting and packaging: AI systems analyze orders and optimize product packaging processes.
• Demand Forecasting: Systems and analyzes sales, seasonality, and trend data to predict future product needs. This helps companies plan purchases in advance and avoid overcrowding in warehouses[2].

Figure 1. Robots for moving goods

Forecasting demand is an important task for optimizing the operation of logistics chains. With the help of artificial intelligence, companies can predict how many products will be needed in the future based on previous sales, market trends, and changes in consumer behavior. For example, retailers can predict an increase in purchases of certain products before holidays or promotions. This allows you to plan purchases more accurately and avoid surpluses or shortages in warehouses.

Logistics is subject to many risks: delays due to weather conditions, transport breakdowns, changes in market conditions, and more. Artificial intelligence can analyze large amounts of data to predict possible risks and suggest measures to minimize them. Thus, artificial intelligence not only transforms logistics, but also opens up new horizons for improving business efficiency[3].

Also, one of the possibilities of using AI in logistics is an automated purchase accounting system using cameras and artificial intelligence, which allows customers to make purchases without the participation of sellers. This technology is used in so-called «stores without sellers» or «cashier-free stores».

 These stores use a variety of components, such as cameras, AI, and behavior recognition algorithms on the ceiling of the store. Cameras are installed in the store where they want to use this technology (for example, 27 cameras in the Standard Market store) that track the movements of customers. The AI analyzes camera data to determine which items the customer has taken, and the system tracks the buyer’s speed and width of steps, the direction of gaze, and the time spent at various items. In more detail, the system works as follows:

Ceiling cameras capture the movements of customers, allowing the system to track which items they take, return, or simply view. This is achieved using machine vision technologies that analyze the speed and direction of movement, as well as the time spent with each product. Next, the AI processes the data received from the cameras to determine the intentions of the buyers. The system is also able to identify potential thefts by analyzing the behavior of customers — long strides and frequent glances towards the exit can signal the intention of theft. If theft is suspected, the system notifies an employee of the store, who then reports the theft incident[4].

Some stores use facial recognition technology to identify customers at the entrance and exit. In other cases, customers use mobile apps with QR codes to enter the store, which allows the system to associate purchases with a specific user.

RFID tags (radio frequency identification) are also used to track goods in real time. These tags allow you to automatically capture information about the goods that the customer takes from the shelves.

These ideas are already being actively implemented in the world, for example, Amazon Go uses a combination of all of the above technologies to create a store where customers can simply pick up goods and leave, and payment is made automatically through the app(Fig.2)[5].

Figure 2. Amazon Go store with a purchase tracking system.

And Standard Cognition has developed a system that uses only ceiling cameras and AI to determine customer behavior without the need for turnstiles or other forms of exit control. The company plans to implement its solutions in more than 50,000 stores over the next five years. This will significantly increase the reach and accessibility of their technologies to retailers around the world.

Standard Market Concept: The company continues to develop the concept of stores where customers can make purchases without having to scan products. The system automatically detects the goods and deducts the payment from the card when leaving the store, which makes the process as convenient as possible for customers.

Camera-based tracking systems and self-service technologies significantly speed up the shopping process in stores. According to various studies and reports, several key aspects can be identified:

Speeding up the payment process: Research shows that 85% of respondents believe that self-service cash registers work faster than traditional cash registers with cashiers. This is because customers can scan and pay for items on their own without having to wait in line. The introduction of self-service systems makes it possible to reduce queues, especially during rush hours. For example, in stores with self-service areas, up to 70% of customers prefer to use such terminals, which significantly increases throughput. Self-service cash registers also take up less space and can be installed in larger numbers compared to traditional cash registers. This allows you to increase the number of simultaneously serviced clients and reduce the waiting time. Although the exact percentage reduction in service time may vary, many retailers report significant increases in service speed due to process automation. For example, the introduction of self-service systems can reduce service time by 30-50% compared to traditional cash registers[6].

However, the implementation of tracking and tracking systems in stores without cash registers may face a number of challenges and problems. For example, systems that use cameras and AI to track customers may raise concerns about the collection and storage of personal data. The need to comply with data protection regulations such as the general data protection regulation (GDPR) can be a major obstacle to implementation. Also, due to the constant monitoring of customers, data leaks or cyber-attacks may occur if there is a vulnerability in the systems, which endangers both customer data and the company’s reputation[7].

Another disadvantage is that the introduction of tracking technologies requires significant financial investments in equipment (cameras, servers) and software. This may not be available to small and medium-sized enterprises, which leaves this niche only for large companies.

Incorrect settings or outdated software may lead to inaccuracies in tracking purchases. Regular software updates are necessary to ensure its reliability. This also leads to the problem of qualified personnel who will be able to manage and control these systems[8].

In the future, it is expected that technology will improve, which will allow the system to more accurately determine the actions of customers and predict their intentions. This includes more detailed behavior tracking, which will help prevent theft and improve security[9].

In a post-pandemic world, when consumers are looking for secure and contactless shopping methods, cashless store technologies are becoming especially relevant. Standard Cognition will adapt to these changes by offering solutions that meet new customer requirements. AI is already being used to optimize delivery routes, manage warehouse operations, and forecast demand, which can significantly improve the efficiency of logistics processes.

Thus, the integration of AI into logistics not only improves the quality of customer service, but also helps to reduce costs and increase the competitiveness of companies in the market. Further development of these technologies is expected in the coming years, which will lead to full automation of many processes in the logistics sector.