Issue 37 Unmanned Systems Technology April/May 2021 Einride next-gen Pod l Battery technology l Dive Technologies AUV-Kit l UGVs insight l Vanguard EFI/ETC vee twins l Icarus Swarms l Transponders l Sonobot 5 l IDEX 2021 report

33 Einride next-gen Pod | Dossier Robert Falck invented a method and system described in a December 2020 US patent application to address these issues in an integrated manner. It does so by taking in and processing multiple parameters and generating an output that amounts to the selection of the best vehicle and route for each delivery task the system is asked to undertake. The system is a controller that could be implemented in software and resident in the cloud, or it could be either partly or wholly implemented in hardware. How it works is perhaps easiest to describe from the point of view of a single delivery job. The basic information the controller needs includes the pick-up and delivery points, the state of charge of the battery packs in all the vehicles in the fleet, all the available routes between the start and end points, and the expected power consumption for each vehicle on each route. The controller then determines the number of required charging cycles for each vehicle on each route and, based on that, selects a combination of a vehicle and a route that can get the job done within a number of charge cycles below a predefined threshold. In theory, taking these factors into account enables the control system to optimise the performance and maximise the service life of each battery pack in the fleet. The controller can also receive other information about the fleet of battery packs, such as the state of health of each pack. It can also be fed with the freight capacity of each Pod, taking into account the weight of cargo currently aboard and that of the cargo to be picked up. Information on the routes can include the number and location of charging stations along each one, and factors such as gradients, traffic, temperature and the current position of each vehicle. All of these can affect the energy consumption calculation, so it is not always the shortest route between the start and end points that is the most energy efficient. Other constraints that can also be factored in include the required pick- up and delivery times and operator availability. This last factor is relevant if the vehicle is semi-autonomous and the route takes it into an area where operator assistance is required, such as past a school. Real freight operations naturally have to accommodate multiple pick-ups and drop-offs in task lists that are constantly changing as jobs are completed and new ones added. When assigning vehicles and routes to jobs, the controller is also designed to ensure that the total number of charge cycles for all the assignments is below a predefined threshold, a capability that uses information on the charge state of individual vehicles to calculate a result at system level. It is intended to increase battery life and reduce maintenance costs still further. Testing started at the same time as system development, as the two always run in parallel with development, according to Hallgren. He adds that all the systems are tested at different levels, including unit testing of software components, simulation testing of system components, simulations of vehicle behaviour, proving ground work and full transportation system validation in cooperation with customers. Working with oat drinks company Oatly, on a pilot integration of electric trucks into routes at its production sites and warehouses in Sweden, taught Einride a great deal about the granular detail of electric freight operation, Hallgren says, enabling the company to update its software and hardware considerations as their partnership has progressed. The same kind of process is going on with customers who have signed up for Pod operations. So far, these include Coca Cola European Partners, logistics specialists DB Schenker and Svenska Retursystem, Electrolux, retailer Lidl Sweden and tyre giant Michelin Group. “All aspects of the Pod are continuously improved as we learn with our customers how to better address their transport needs,” Hallgren says. “Pod development over the coming years will focus on extending the system to larger and larger ODDs. These capabilities will evolve such that all levels from AET 1 to AET 4 will be addressed by 2024.” Unmanned Systems Technology | April/May 2021 Schematic of the transport planning system, which takes into account factors including the required delivery times and the locations of charging stations when allocating a vehicle to a job