Author: Cai Yuli Zane Tsai, product director of embedded platform and module division of Linghua technology platform product center

Automation makes use of technology to enable human beings to complete more tasks. In the field of logistics, automation has great potential and its benefits are obvious, especially when the operation mode changes greatly or the demand is increasing. Expanding the operation scale usually requires additional employees, and these employees are usually unable to work immediately, especially when other industries have similar needs. How to respond quickly to market fluctuations requires rapid action and other additional capabilities in the whole process of operation.

Logistics automation can quickly increase capacity according to changes in demand. After promoting logistics automation to a strategic position, it can not only improve productivity, but also reduce human errors and improve work efficiency. With appropriate logistics automation software, hardware and platform resources, even in the period of low demand, the impact on the operation cost is relatively small, which is far lower than the cost required to maintain a large number of manpower. As demand increases, operational capabilities are ready and can be started quickly. Although these methods can bring the required flexibility to logistics companies and quickly respond to changes in demand, there are still opportunities to do more.

Artificial intelligence will amplify the influence of logistics automation

Introducing artificial intelligence into logistics automation will greatly enhance the influence of artificial intelligence. AI can reduce errors in common semi skilled tasks, such as classifying and sorting products. The use of autonomous mobile robot AMR can improve the efficiency of package delivery, including the most expensive delivery of the last kilometer. Artificial intelligence helps the autonomous mobile robot AMR to plan routes and identify features, such as people, obstacles, delivery portals and doorways.

Integrating logistics automation into any environment will bring some challenges. It can be as simple as a power conveyor instead of a repetitive process, or as complex as introducing a cooperative autonomous robot into the workplace. When artificial intelligence is added to the process of automation and integration, the challenges will become more complex, but the benefits will increase.

The efficiency of each automation element will increase with the interconnection between solutions and a deeper understanding of other stages in the process. Placing AI near devices that generate data and take action is called edge AI. The adoption of edge AI is redefining logistics automation.

The development of edge artificial intelligence is extremely rapid, and its application is not limited to logistics automation. The benefits of placing AI at the edge of the network must be balanced with the availability of resources, such as power, environmental operating conditions, logistics location and available space.

Reasoning at the edge

Edge computing makes computing and data more closely combined. In traditional Internet of things applications, most of the data is sent to the (cloud) server through the network, where the data is processed, and then the results are returned to the edge of the network (such as the physical device at the edge). Only cloud computing introduces the consideration of delay, which is unacceptable for time sensitive systems. Here is an example of the role of edge computing. In the sorting process, capturing and processing the image data of parcels locally can make the logistics automation system respond in just 0.2 seconds. The network delay in this part of the system will make the sorting process slower, but edge computing can eliminate this potential bottleneck.

Although edge computing makes the calculation closer to the data, the introduction of artificial intelligence to the edge side can make the process more flexible and less error prone. Similarly, the logistics of the last kilometer largely depends on manpower, but the autonomous mobile robot AMR using edge artificial intelligence can improve this situation.

The introduction of artificial intelligence will have a significant impact on the hardware and software used in logistics automation, and there are more and more potential solutions. Generally, the solution for training artificial intelligence models is not suitable for deploying models on the edge of the network. The processing resources for training are designed for the server, and its demand for resources such as energy consumption and memory is almost unlimited. At the edge, energy consumption and memory are limited.

Heterogeneous trends

In terms of hardware, large multi-core processors are not suitable for edge AI applications. Instead, developers are deploying optimized heterogeneous hardware solutions specifically for edge AI. This scheme includes CPU and GPU. Of course, it can also be extended to ASIC, MCU and FPGA. Some architectures (such as GPU) are good at parallel processing, while others (such as CPU) are better at sequential processing. Today, no single architecture can really provide the best solution for artificial intelligence applications. The overall trend is to configure the entire system with hardware that provides the best solution, rather than using multiple instances of the same architecture.

This trend points to heterogeneity, where there are many hardware processing solutions with different architectures that work together through configuration, rather than the same architecture using multiple devices (all devices are based on the same processor). The ability to introduce the right solution for any given task or integrate multiple tasks on a specific device can provide greater scalable performance and optimized performance per watt and / or per dollar.

Moving from homogeneous systems to heterogeneous processing requires a huge solution ecosystem and a mature ability to configure these solutions at the hardware and software levels. That’s why we need to work with a supplier that has the ability to establish partnerships with all chip suppliers, because this supplier can provide solutions for edge computing and work with customers to develop systems with scalability and flexibility.

In addition, these solutions use general open source technologies such as Linux and professional technologies such as robot operating system ros2. In fact, more and more open source resources are being developed to support logistics and edge AI. From this point of view, there is no single “right” software solution, and so is the hardware platform running the software.

Using modular method to construct edge computing

In order to improve flexibility and reduce being bound by suppliers, Linghua technology has developed a modular method at the hardware level, which can make the hardware configuration in any solution more flexible. In fact, modularity at the hardware level allows engineers to change any part of the system hardware, such as the processor, without system wide interruption.

When deploying new technologies such as edge artificial intelligence, the ability to “upgrade” the underlying platform (whether software, processor, etc.) is particularly important. Each new generation of processor and module technology usually provides a better power / performance balance for the reasoning engine at the edge of the network, so it can quickly use these performance and power gain to minimize the interruption of the whole logistics automation system, and the edge artificial intelligence system design is also an obvious advantage.

By using micro service architecture and docker container technology, modularization in hardware is extended to software. If a more optimized processor solution is available, even if it comes from different manufacturers, the software utilization processor is modular and can replace the previous processor without changing the rest of the system. Software containers also provide a simple and powerful way to add new functionality to run in edge AI.

The software in the container can also be modular. The edge vision Analytics (EVA) SDK (software development kit) of Linghua technology for artificial intelligence vision products is a typical example. Based on GStreamer, the platform focuses on the basic functions required to build artificial intelligence vision pipeline. At each stage of the artificial intelligence vision pipeline, ready-made open-source plug-ins (including their own modules) are used to simplify the development of the pipeline. These plug-ins include image capture and processing, artificial intelligence reasoning, post-processing and analysis.

The modular and container approach of hardware and software minimizes the risk of being bound by suppliers, which means that the solution does not depend on any specific platform. It also increases the abstraction between platform and application, making it easier for end users to develop their own applications that do not depend on any platform.

We simplify the upgrade process through a database that characterizes components when they are available. Using this database, engineers can select suitable products to achieve a perfect balance between reasoning performance and system resources.

One of the most important requirements of logistics automation is to respond in real time. Therefore, it is very important to cooperate with a supplier who has rich experience in software and hardware combined development system and can meet the application requirements. Linghua technology’s approach is to use modules that can be integrated with professional third-party technologies, such as lidar sensors.

conclusion

Deploying edge AI in logistics automation does not need to replace the whole system. First, you need to evaluate the workspace and identify the stages that can really benefit from AI. The main goal is to reduce operating expenses and improve efficiency, especially in the period of labor shortage, so as to cope with the increase of demand.

More and more technology companies are committed to developing artificial intelligence solutions, but most companies usually only focus on cloud computing rather than edge computing. On the edge side, its operating conditions are different, resources may be limited, and even a private network may be required.

Automation will continue to grow and expand in logistics operations through the use of technologies such as artificial intelligence. These system solutions need to be specially designed to meet the harsh operating environment, which is very different from the requirements of cloud or data center. We use a modular approach to solve this problem, which provides a highly competitive solution, shorter development cycle and flexible platform.

Leave a Reply

Your email address will not be published. Required fields are marked *