5-stage framework for Evocargo service deployment

How we deploy autonomous vehicles to improve long-established logistic processes at factory yards

Robots, self-driving vehicles, and other autonomous systems are ever more visible in today’s workplace. All credit to their inventors. But we often forget that designing and building a robot is only half the battle. The other half is ensuring that the robot can be deployed in a working environment to collaborate efficiently and safely with other machines and with the people who work there. This applies to all kinds of automated devices, from a robot vacuum cleaner to an industrial robot or an autonomous vehicle (AV).

This challenge seems a tough call. Future use of AVs on city streets will require AV-welcoming legislation, greater 5G coverage for connected vehicles (V2X), powerful servers and operational centers to store and process the huge data stream from the vehicles, not to mention changes in urban design from new signage to smart parking lots and much more. Those changes will require years of research, substantial investments, and a revolution in how people use cars.

By contrast, factory yards, warehouses, port zones, etc., have restricted access and are usually protected by a fence. They put fewer legal constraints on AV operation and they are more predictable for the autonomous vehicles that are put to work there. As cargo has to be transported non-stop, sometimes under harsh weather conditions, there is a demand for automation and modernization in this area.

Evocargo pioneered yard logistics several years ago, and our AVs are in commercial operation, carrying cargo at a range of industrial sites, from metallurgy plants to logistics hubs.

Most of the customers we work with have long-established processes and routes around their sites. They need us to introduce automation gradually and “silently”, without obstructing the work of other vehicles and people in the existing infrastructure.

We provide:

• the AVs themselves, electric autonomous trucks (Level 4 by SAE standard J3016_202104), fully designed and assembled by us. By using alternative energy sources and solving the problem of labor shortages the AVs reduce our customers’ costs and increase efficiency at their sites;
• a freight management system, which allows monitoring the rides and resolving difficult situations en route in real time.

We have our own multi-stage procedure for deployment of our AVs at each new site. It is a procedure that we designed from scratch when we started work several years ago and that we have been improving ever since.

Deployment stages for Evocargo service

We work through five main stages to match our service and AVs’ capabilities to the requirements and site specifics of each customer.

Stage 1. Initial assessment

Our operations team arrives at the site to get an overview of the infrastructure. They check:

• electricity supply for charging stations;
• the parking lot;
• internet stability along all potential routes;
• a workplace for the fleet operator (if required on-site);
• potential bottlenecks on the route, which may slow down logistics or cause safety risks.

Some sites have new roadways and straightforward routes. Others lack full network coverage and their routes cross rail tracks. Our engineers take notes of such specifics and offer changes if needed.

Stage 2. Vehicle delivery

When all the technical requirements have been settled with the customer, the operations team prepares the requested number of vehicles, runs complex checks on the hardware and software, and accompanies delivery of the vehicles to the site.

Stage 3. Digital twin

The operations team works on site to create a rich and informative map (the digital twin) of the space. They use sensory data obtained using:

• quadcopters;
• a set of sensors installed on an ordinary car, imitating the sensor configuration of an Evocargo AV;
• or an Evocargo AV that drives around the area and captures the area with its cameras, lidars, and other sensors.

The map is annotated with site-specific features: loading docks, routes, charging stations, parking lots. If there are no signs and road markings at the site, the operations team marks the best places for U-turns or speed limits on the digital map to compensate the lack of physical navigation.

All these data will be used by the vehicle’s autonomous system and by the freight management system. Moreover, the digital twin of the site can be used as a simulated environment where we test the autonomy features of our vehicles before exploitation.

Stage 4. Launching freight management

Our freight management system:

• shows the customer how the autonomous fleet operates;
• enables the customer to manage orders;
• serves as a powerful tool for Evocargo operators to detect and resolve issues in a timely manner, remotely dispatch vehicles and manage routes, analyze all rides, highlight bottlenecks, and more.

All this functionality is packed into a single cloud service that can work on any device and anywhere (at the customer’s location or remotely at Evocargo’s own facilities).

To launch the service, our engineer creates a register of vehicles delivered to the customer and assigns roles for maintaining the service and resolving technical issues (usually the responsibility of an Evocargo specialist) and for viewing rides info (usually the responsibility of the customer’s own staff).

Stage 5. Final adjustments

Prior to the first ride, the operations team checks cameras and lidar calibration on the AVs, tests navigation, localization and safety systems, and carries out other essential checks.

The Evocargo team then supervises the first few rides to detect and eliminate any inconsistencies in the map or to optimize routes if necessary. We check that all maneuvers are smooth and that the vehicles approach and halt at loading docks (sideways or backwards) with accuracy of 10 cm or better.

Stages 3 to 5 can take from 3 days to 2 weeks. After the AVs have driven all the routes successfully in fully automated mode, the deployment work is complete and the vehicles are ready to enter service.

Regular operation

Once the vehicles have started regular automated service, only one Evocargo operator is needed to monitor and maintain a fleet of up to 10 vehicles. The operator makes sure the vehicles get the latest software updates, deals with any infrastructure changes or new requirements of the customer, remotely resolves issues on routes and detects bottlenecks.

Our operations team

The Evocargo operations team saves customers the trouble of day-to-day AV fleet management. Our engineers:

• know how the vehicle is designed;
• understand the logic of all-electric powertrain and other cutting-edge technologies;
• understand the customer’s needs;
• are familiar with Linux and some scripting language such as Bash or Python;
• know their way around toolbox in order to fix things on-site;
• are committed to making the system perform even better for the customer.

Evocargo engineers have a wealth of experience matching Evocargo AVs to the realities of our customers. That experience is ready and waiting to make logistics business work better.