Industrial wireless network provider Rajant Corporation is working on a project with autonomous vehicle specialist DGWorld to design a backbone solution for DGWorld’s selfdriving container terminal trucks used in port operations.
Ahead of TOC Asia last month, UAE-based DGWorld and Rajant announced that they had entered into a strategic partnership to work together on products and solutions using Artificial Intelligence (AI), automation and self-driving systems. DGWorld could not provide extensive details at this point, but the companies are working on a project to convert around 150 existing terminal tractors at a large seaport in the Middle East to operate completely autonomously, along with other projects, such as driverless taxis, shuttle busses and other forms of public transport.
DGWorld’s autonomous technologies combine LiDAR and camera data to classify and track objects. This helps identify people, roadway details such as lane lines, and other vehicles encountered along the tractor’s self-driven pathway. “In a terminal tractor application, this would mean the machines would operate around the ever-changing dynamics of a port without human intervention,” said DGWorld.
“Such a mobile application requires continuous connectivity for never-fail reliability from the wireless network,” the firm emphasised. This is a real challenge in a port application. The AGVs used at automated terminals today follow transponders in the ground, and only send and receive limited data, such as work instructions, via WiFi. By contrast, what is being contemplated with this application from Rajant Corporation is a much more data-intensive wirelessonly operation, including the ability to monitor machine location and status in real time, as well as connect any terminal tractor with a remote operator to drive the machine semi-autonomously, if necessary. “Continuous
wireless communication is mandatory”, stressed DGWorld.
Speaking with WorldCargo News at TOC Asia, Geoff Smith, Rajant’s EVP global sales and marketing, said: “It is our patented Kinetic Mesh V2X technology that is well suited to the task. Unlike WiFi networks that rely on a static, overhead antenna to blanket a terminal with coverage, Rajant’s Kinetic Mesh V2X network moves with each terminal tractor because it is equipped with one of Rajant’s BreadCrumb wireless nodes connecting it to our mesh network, enabling uninterrupted operation of the vehicle’s sophisticated control system. These BreadCrumbs use up to four radios to communicate with the network and with each other. If at any point a node cannot connect directly to the V2X network, it connects through other V2V nodes, using Rajant’s proprietary InstaMesh software to adapt to a constantly changing environment.”
Kinetic Mesh is designed to support high throughput rates for data, voice and video signals at up to 300 Mbps, depending on the signal strength. The actual data throughput rate required depends on the application, and in particular how much intelligence is on board the vehicle. For the terminal tractor automation project, the TOS will send work orders and suggested routes to the tractors, but the execution of the work orders is managed by onboard software.
“What is more important than high data throughput is this application’s ability to have continuous connectivity without interruption and very low latency,” continued Smith. “No technology exists today that can see through metal, so the only way to achieve continuous communication in a port terminal is to connect around the container stacks. This requires V2X communication, which allows vehicle-to-infrastructure and vehicle-to-vehicle connectivity. The autonomous tractors will have a dedicated network and use our multi-radio V2X technology to manage interference from other networks.”
Interference is a particular problem around vessels, which often have their own WiFi system and generate disruption. Kinetic Mesh can hold multiple connections simultaneously over multiple frequencies, so signals can be easily diverted if one becomes blocked. Moreover, the system uses machine learning to identify and route communications via the next-best available option, overcoming any obstruction instantaneously and autonomously without outside intervention.
Low latency is also a requirement to support the type of remote operation the terminal operator wants when implementing any level of semi-autonomous operation. Early testing by DGWorld and Rajant identified that a remote control desk with joysticks, typically used for cranes, is not ideal for terminal tractors. Instead, the terminal operator wants a more tactile system with a steering wheel and two pedals, with force feedback for the brake pedal. Kinetic Mesh can support this with sub-millisecond latency. Ton Bijlaart, managing director at DGWorld, said he is confident Rajant’s Kinetic Mesh can support the deployment of its autonomous vehicles in container terminals.