Self-organising logistics

Pilot Autonomous Algorithms

If intelligent individual trucks can make decisions and are granted the autonomy to do so, can these trucks create an effective and efficient logistics system? In Autonomous Algorithms we seek answers to these questions, work towards a real-life pilot and take a first step towards self-organization.

Self-organizing shipments are routing themselves through a global distribution network. Welcome to 2050 logistics. Could this be the solution to the increasing demands on freight transport?

Many of us drive to work every morning. Few of us carpool with others in the area. An example of self-organization, effective, but logistically very inefficient.

Watch TNO's video about Autonomous algorithms

Why is it that we want to decide when to get in the car?

Possibly, because we possess information that a transportplanner lacks. An appointment with a customer, picking up the kids from daycare, a missing toothbrush. Local information that is unavailable to a central planner.

Maybe the matching process is just not possible. How much effort would it take to inform hundreds of people in the neighborhood about our ever-changing wishes and time schedules? What is the chance that someone else wants to move in the same direction at the same time?

No wonder logistics organizes this differently. Transportorders are collected, a cutoff-time determined, and the fleet is planned. The goal: Deliver cargo at the customer’s site at exactly the right time with minimal cost.

What if matching would not take excessive time? If it would save money? If it would halve your CO2-footprint? And what if you could leave whenever you wanted?

If not a human, can a truck do this? Organize a logistics service autonomously?

Imagine a truck that calculates the costs of a transport, checks whether there is time in the schedule, whether there is a driver available, and predicts how fast and effective it thinks the transport may be executed.

In that scenario an incoming transport order may be matched with the truck that communicates the most (system-)efficient prediction. If that truck monitors the effectivity of the transport it can use this data to improve its future predictions.

Welcome to "Autonomous Algorithms"

Just like a people, a truck has access to local information. The truck knows whether it is delayed in a traffic jam, keeps maintenance into account and can communicate with other trucks to take over cargo whenever needed.

All of this information can be processed in ever bigger and more complex central systems. But if we do this, can we still rapidly re-plan if a single truck experiences delay? What if that system fails? Can we create a system that is more effective than the current system? And can communication between individual trucks ensure that we don’t lose the efficiency lost in our daily commute?

In Autonomous Algorithms we developed a proof-of-concept to validate whether a decontrol system is a technically feasible. In the simulation environment trucks autonomously organize their transport planning and execution.

Together with DHL Global Forwarding and Van Berkel Logistics we are working towards a pilot. In this pilot we will use the software in an actual real-life scenario to seek answer to the questions above.

A first step towards self-organization.

Learn more

Read the whitepaper 'Exploring Possibilities of Letting Vehicles Plan and Organise Transportation Themselves'

Get inspired

1 resultaat

First steps towards self-organising logistics

Automation and robotization are on the rise. These technologies are set to change the logistics sector drastically. Read more.