Autonomous forklifts, a type of AGV and AMR, have transformed the future of warehousing, manufacturing, and intralogistics with advanced autonomy. This advanced AI-powered robotic technology has made significant impacts in enhancing workplace safety, workflow efficiency, and scalability. In this case study, we’ll discover how an automation robotics company leverages Premio’s AI Edge Inference Computer to drive their autonomous forklift solution.
Challenges:
- Sought out a reliable edge computing solution to process edge AI workloads in real-time
- Sufficient and compatible IoT connectivity to consolidate the on-board electronics and sensors
- Low-profile design to be retrofitted into the already space-constrained forklift chassis
- Operational reliability for in-vehicle deployment with varying warehouse conditions
Solution:
- Premio’s RCO-6000 Series AI Edge Inference Computer (RCO-6000-CML-2)
- EDGEBoost Node with dedicated GPU support for streamlined edge AI workflows
- Modular EDGEBoost I/O configured to support dual 10GbE LAN for high-fidelity IoT sensors
- Rich on-board connectivity for critical electronics
- In-vehicle durability to withstand rigorous warehouse environments
- World Class Certifications: UL, FCC, CE
Benefits:
- Direct partnership with manufacturer
- Dedicated engineering support
- ISO certified facility
Company Background
Founded in 2017 and backed by major players in the automotive and retail space, this automation robotics company honed their efforts in bringing intelligent automation to the logistics sector. Their intelligent robotic forklift solution is designed to boost productivity by addressing warehousing bottlenecks and material flow inefficiencies, while creating a safer workplace environment. By leveraging advanced technologies such as computer vision, machine learning, and edge AI, their unmanned solution can navigate complex warehouse layouts, avoid obstacles, and perform precise tasks. The robotics company’s mission to drive AI-powered autonomy will make tremendous advancement strides in material and pallet handling processes.
The Challenge
The automation robotics company encountered ongoing issues and limitations with their previous solution. The form factor significantly restricted its integration options within the already space-constrained forklift. This resulted in non-compliance with an OSHA standard, which requires a proper field of view if there is a circumstance where a driver is needed to manually operate the forklift. In search of a more reliable ruggedized edge computing solution for their new autonomous forklift model, they compiled a detailed checklist of deployment requirements that needed to be addressed.
In-Vehicle Deployment Reliability
Due to the deployment environment conditions and space limitations of the forklift chassis, the robotics company needed an edge AI computing solution with a minimal footprint and was durable to withstand in-vehicle deployments. Their previous edge computer was too large, significantly limiting the placement options into the forklift. With varying warehouse environments, the edge computer needs to withstand frequent shock and vibration, varying temperatures, and have a maintenance-free design for prolonged reliability.
AI Inferencing Performance for Computer Vision
A major requirement was a need for the industrial computer to streamline their autonomous AI software without performance bottlenecks or throttling. Since the computer is deployed onto a forklift, considerations for power consumption, thermal management, and overall durability were crucial in selecting the proper performance accelerator. Although a high-performance GPU can seamlessly drive the AI inferencing capabilities, it would not be suitable within a forklift due to key limitations of space, thermal regulation, and even IoT sensor connectivity. The robotics company required expertise on integrating an optimized performance accelerator that can balance both AI performance and edge computing ruggedness.
High-Speed IoT Connectivity & Throughput
The autonomous forklift utilizes various IoT sensors and devices to precisely map and navigate through the warehouse environment. These IoT sensors have different connection types, from legacy COM to high-speed USB and LAN ports. Additionally, the forklift incorporates high-bandwidth LiDAR sensors for extremely accurate mapping and requires a 10GbE bandwidth connection. The industrial computer needed to provide the necessary amount of I/O support and be easily compatible with the mixture of multiple connections to a variety of sensors.
The Solution
Real-Time Edge AI Performance with EDGEBoost Nodes
Enabling the autonomous forklift’s edge AI software requires a balance in performance, size, and durability. By leveraging modular EDGEBoost Node technology, the RCO-6000 Series is configurable with a dedicated low-profile GPU to streamline edge AI workloads.
Premio chose the RTX A2000 due to its AI-oriented GPU acceleration performance, low-profile form factor, and power efficiency. This combination provides the RCO-6000 Series an ample amount of parallel processing power to handle edge AI workloads while maintaining durability for in-vehicle deployment.
| NVIDIA RTX A2000: | |
|---|---|
| GPU Memory | 12GB GDDR6 with ECC |
| Max Power Consumption | 70W |
| Form Factor | Low-profile Dual Slot |
Seamless IoT Consolidation with EDGEBoost I/O Technology
The RCO-6000 Series simultaneously provides rich IoT-centric I/O on-board and offers I/O flexibility with modular EDGEBoost I/O technology. The on-board I/O features a variety of IoT-specific connectivity, from the standard COM, USB, and LAN to isolated DIO. For additional customization, the RCO-6000 Series supports dual EDGEBoost I/O brackets, allowing for seamless configurability with additional I/O to meet high-specification requirements. In this case, the autonomous forklift utilizes a LiDAR sensor that requires a 10GbE connection, a key requirement for the autonomous forklift. With a standardized and modular design, Premio configured the RCO-6000 Series with a dual 10GbE EDGEBoost I/O module to meet this requirement with ease.
Long-Term Product Support Lifecycle
Alongside AI performance, Premio configured the RCO-6000 Series with a 10th Gen Intel TE processor. By opting for the TE model, Premio can provide long-term deployment consistency since this line of Intel processors is under Intel’s 10-year embedded support lifecycle roadmap. This allows the robotics company to integrate a repeatable edge AI computing solution without sudden design changes that may conflict with the overall solution.
In-Vehicle Durability & Reliability
The RCO-6000 Series is purpose-built with rugged features to withstand industrial deployments in extreme conditions and environments for 24/7 in-vehicle use cases. Its fanless and cableless design approach prevents the ingress of dust and debris while enabling a list of other crucial features:
- Wide Operating Temperature Range (-25°C to 70°C)
- Shock & Vibration Resistance (MIL-STD-810G Compliance)
- Wide Power Input Range (9~48VDC)
- Power Protection (OCP, OVP, RPP)
- Power Ignition Management
- CAN Bus Support
- World Class Certifications: UL, FCC, CE
Forklift Form Factor Compatibility
The RCO-6000 Series had the smallest footprint when compared to other edge computing options. It was retrofitted into the forklift chassis with ease, did not interfere with other electronics or protruded outwards preventing the enclosure door from closing. This now allowed the robotics company to meet an OSHA compliance standard that was unachievable in the previous solution.
The Benefits
Premio not only provided a ruggedized edge computing solution that met all the robotics company’s requirements but also offered additional benefits as an official partner. As the direct manufacturer, Premio provides value-added support through its world class engineering, supply chain management, and turnkey manufacturing services typically unavailable through oversea competitors with little presence in the United States of America.
Direct Manufacturer Partnership
By partnering directly with the manufacturer, Premio brings its 35+ years of edge computing expertise and support to the automation robotics company. From supply chain management to technical engineering support, the robotics company was able to rely on Premio for knowledgeable computing experience and standard solutions, eliminating the need to allocate expensive resources and market delays with other solutions.
Proactive Sales & Engineering Support
Premio’s sales and engineering teams offered exceptional support to ensure that the rugged edge computer exceeded the expectation of the client’s standards. When the automation robotics company needed to make software adjustments with a custom BIOS, Premio’s BIOS engineers quickly collaborated with them to configure and optimize the AI software with turnkey ease.
Localized Supply Chain Management
Strategically located in Greater Los Angeles area, Premio’s dedicated supply chain management team is experienced in navigating supply chain disruptions and making effective decisions to ensure a just-in-time delivery approach for clients OEM computing solutions.