An Autonomous Tractor Partnership
Terberg is Working with ZF to Develop an Autonomous Terminal Tractor
TAGS: Terberg, Autotug, ZF
Terberg has revealed another important step in the development of the AutoTUG automated terminal tractor that it first announced at TOC Europe in 2016. The company is now working with driveline, chassis and safety technology specialist ZF, to develop an autonomous driving system for AutoTUG that will allow it to operate in mixed traffic terminals, where automated tractors will need to share the traffic lanes with road trucks. This autonomous terminal tractor will use ZF’s ProAI system to process sensor information in real time.
Moving forward with autonomous driving is important to match the demands of the market. Terberg initially developed AutoTUG using fusion layers to combine various techniques that fulfil the same function, such as a combination of RTK GPS and the RFID transponder system from BTG Positioning Systems for navigation. In line with this technology, the first applications on the AutoTUG product roadmap were fully automated systems, where the machines would operate in closed zones. It was planned that mixed terminals that would require an autonomous driving system would follow at a later date.
Rob van Hove, Terberg managing director, said there is tremendous interest in AutoTUG, but “everyone wants to use it autonomously from the start”. At new terminals, where there is the opportunity to design the layout to separate manned and unmanned equipment, terminal operators currently prefer to use automated machines, like AGVs. But for existing RTG terminals, where shared traffic flow is required, the market wants autonomous solutions.
AutoTUG was designed from the outset to be flexible with regard to navigation and collision-avoidance systems, communication systems and autonomous vehicle technology. Terberg developed the automation architecture using three so-called function layers. The first comprises the lower level sensors and actuators that control the AutoTUG’s steering and fifth wheel, the second is the navigation system, and the third is a vision layer for collision-avoidance and autonomous control. The systems at each different level can easily be equipped with new technology or other applications as required. This flexibility has allowed Terberg and ZF to take the AutoTUG developed for the initial tests in the Netherlands, and install a new autonomous vehicle system from ZF.
ZF is developing an autonomous vehicle system that can be adapted to several applications, like cars, trucks or off-road vehicles. Its technology includes cameras, radars and laser-based sensors for navigation, collision-avoidance and pedestriandetection in autonomous vehicle applications. To achieve autonomous driving, a central computer called ZF ProAI processes all the sensor information in real time, makes decisions, and calculates a route.
Terberg and ZF are now working together to integrate ZF ProAI as a modular layer in the AutoTUG system, specifically for applications that require a mix of manned and autonomous vehicles, including a standard RTG terminal.
The partnership with ZF does not affect Terberg’s agreement with Konecranes to make AutoTUG available through the latter firm’s network, alongside Konecranes’ other automated equipment. Terberg is working with ZF, said van Hove, to develop the autonomous capability of the AutoTUG, and the machine itself will remain compatible with Konecranes’ other equipment and its TEAMS Equipment Control System (ECS).
Integrating ZF’s ProAI system changes the relationship between the navigation and the ECS system. As van Hove explained, AutoTUG was initially developed to follow a given path defined by the ECS. With the yard divided into a grid, using RFID transponders, the ECS controls the vehicle by releasing segments or zones of the yard to the machine incrementally. The ECS knows where all the equipment is in real time, and can avoid any situations where vehicles operate on the same path at the same time, and the AutoTUGs are reliant on the ECS for instructions.
By adding ZF ProAI as a modular layer, van Hove continued, Terberg is looking to take the next step and introduce more intelligence to the vehicle, so it can resolve issues on board, without the complexity of all decisions being made in the ECS.
This is a complex challenge, but it is necessary, in order to meet the goal of autonomous driving within an RTG operation, or a distribution centre environment. Interest in automated tractors remains very high, particularly in the Middle East, where there are lots of large RTG terminals that rely heavily on expatriate labour.
Terberg has moved mixed environments and autonomous driving up the priority list on its roadmap, and is targeting having the development completed by 2021, but van Hove cautioned that Terberg and ZF are not yet in a position to say when the technology will be available for full-scale commercial deployment. Much depends on whether terminal operators re prepared to commit to a project and partner on the development work. The complexity, however, has not dampened Terberg’s commitment to bringing an autonomous tractor to the