Making warehouse more efficient by Artificial Intelligence
With the crazy booming of online shopping in recent years, the workload of logistics workers has been overwhelming. As labor costs rise and labor shortages continue, and the working environment in warehouses poses a safety concern for workers, all these factors catalyze automation in the labor-intensive warehouse.
Nowadays Artificial Intelligence (AI) and machine vision are gradually being applied to warehouses; it heightens efficiency and ensures safety, and workers can get rid of manual work for more valuable roles. According to the report from LogisticsIQ, warehouse automation is expected to be worth an estimated USD30 billion by 2026, growing at a CAGR of 14% during the forecast period.
Using industrial PC, computer vision, and edge AI systems in various applications can help construct a full-scale automated environment.
Robotic arms take the tasks via machine vision
We can make the most use of robotic arms to lower the workload of repetitive and labor-intensive tasks like picking and sorting goods in warehouses for workers. It’s where the machine learning and visuals kick in. Through machine learning and vision, robotic arms can learn to identify countless items by shape, size, and color, then accurately grab and place them to the designated places for transport to accelerate the logistics procedure.
Axiomtek’s IPC950 would be a competitive choice. Its high performance comes from the 11th Gen Intel® Core® i7/i5/i3 or Celeron® 6305E processors and up to 64GB DDR4 SO-DIMM that the IPC950 can handle computing and operational control at meanwhile without an additional accelerator card. It features front-accessible I/Os for easy access when it is wall-mounted or placed inside a cabinet. Four USB 3.1 Gen2 are ready to connect to multiple high-resolution cameras. Also, it supports an additional PCIe x4 slot for motion control and frame grabber cards to enable the IPC950 to detect the incoming item, identify it then grab and place it in the box for delivery. What’s more, the 2.5G LAN port makes all the work smoothly without latency.
More native and expandable I/Os, such as LAN, 8-CH DIO, COM, DisplayPort and a flexible I/O module slot for various sensors, actuators, motors, and displays, are ready to meet the needs of various robotics applications.
AMR fleet carries the heavy load for workers
Autonomous Mobile Robots (AMRs) are making a difference in warehouses; they can deliver heavy loads of items to designated areas for storage or deliver goods to workers for pickup. The AMRs have to plan the best route and navigate the warehouse floor on their own, avoiding obstacles and communicating with the fleet at the same time. For this application, it is essential to choose a high-performance, power-saving, and compact rugged embedded computer with diverse I/O.
AMRs not only lower labor costs but can also work with people, then optimize efficiency. AMRs’ importance is self-explanatory; the market has been expanding according to ABI Research, the global AMR market will grow from USD 800 million in 2020 to a mind-blowing USD 49 billion by 2030.
Accordingly, Axiomtek proposes the AIE900-XNX, our latest edge AI embedded system with a rugged and low-power design. It’s powered by the NVIDIA® Jetson Xavier™ NX, providing extraordinary computing capability for AI mapping and automated operation. Measuring 239 x 185.3 x 79.4 mm (WxDxH), the AIE900-XNX fits easily into an AMR; moreover, it has rich I/O interfaces, including COM, PoE, LAN, USB 3.1, HDMI, CANbus and 8-CH DIO to connect with 3D LiDAR sensors, 3D cameras, displays, and other end devices. There is an M.2 B-key 3042/3050 slot with a SIM slot for a 5G module and a PCI Express Mini Card slot for a Wi-Fi/LTE module. Five antennas are available.
To meet application challenges in a wide range of industries, the AIE900-XNX offers -30°C to +60°C wide temperature support with 10% - 95%, non-condensing relative humidity tolerance and offers 3Grms vibration resistance. The system is powered by a 9-36 VDC terminal block input with optional ignition power control.
Safety monitoring system to take care of the workers
According to US Occupational Safety and Health Administration (OSHA), in the 10-year span between 2011 to 2021, forklift-related incidents due to operator error or improper use have caused 900 death and an annual average of 7000 injuries. To prioritize safety in such a fast-paced working environment, AI can give you a hand.
As sensors are installed on forklifts, every operator’s driving pattern can be tracked. These data can be further analyzed to enhance safety and be utilized for future preventive training. Personal protective equipment like helmets is also essential for workers. Via machine learning and vision, the workers without wearing suitable equipment can automatically be detected with IP cameras and notified.
Axiomtek proposes the AIE100-T2NX to leverage AI inference technology to perform safety monitoring. The AIE100-T2NX is powered by the NVIDIA® Jetson™ TX2 NX which has Pascal™ architecture GPU with a 256-core to deliver exceptional graphics processing. To easily install and deploy an IP camera, this compact AI-enabled system is equipped with a 15W GbE PoE port for both power and video transmission. In addition, the AIE100-T2NX has one PCI Express Mini Card slot and one SIM slot for wireless network connection to transmit data to central management.
The AIE100-T2NX can operate under a wide range of temperatures from- 30°C to +60°C and withstand vibration up to 3 Grms. With wall mount brackets, fanless design and waterproof IP42-rated cover kit which can help you easily install in a rugged environment.
Machine vision for speedy barcode reading and freight dimensioning
Reading barcodes of goods in the blink of an eye strengthens traceability and makes sure the last mile delivery is reliable. We need more than the utility of the code reading algorithm. As countless boxes and pallets flock into the warehouse, a fast machine vision system to read is fundamental. When the incoming good triggers the sensor, the reader should be able to capture the code within a second.
Machine vision also influences your revenue. It can be deployed for dimension measuring. Regardless of the size and shape, even the boxes are in motion, the length, width, and height will be identified to calculate the shipping fee and sort for delivery. To ensure the quality of the shipment and avoid complaints, machine vision and AI can also assist with defect detection. Once a box or barcode is found damaged, the staff can resolve the issue as soon as possible as the central management is notified by the edge devices. Therefore, we can avoid missing freight and damage claims from customers. It can be accomplished with a high-performance system that can run sophisticated algorithms.
Powered by advanced AI, Axiomtek’s AIE900-902-FL is excelled in graphics processing, it becomes a perfect fit with 3D camera for fast dimensioning and 1D/2D barcode scanner for speedy barcode reading. The AIE900-902-FL is powered by the NVIDIA® Jetson AGX Xavier™ platform, featuring the 512-core NVIDIA Volta™ GPU GPU with 64 Tensor Cores, capable of AI processing speeds up to 32 TOPS. The AIE900-902-FL is equipped with two USB 3.1 Gen2 ports and two USB 3.1 Gen1 ports for hooking up 3D cameras or barcode scanners. The high-performance AI system further comes with four PoE ports for additional cameras.
In addition to the AGX Xavier’s 32GB LPDDR4 and 32GB eMMC, the AIE900-902-FL is equipped with an M.2 M-key 2280 slot with PCIe x4 for optional NVMe SSD. There is also a MicroSD slot and a swappable external 2.5-inch SATA bay to meet different data storage needs.