Autonomous Forklifts use Advanced Computer Vision and Artificial Intelligence

A fleet of autonomous self-optimizing forklift trucks could work alongside human co-workers and automatically adapt to changing work demands.

Pictured L-R with the forklift prototypes is Professor Tom Duckett, Principal Investigator for the University of Lincoln's ILIAD team, Dr Martin Magnusson from Örebro University, and the rest of the team at Lincoln, Dr Grzegorz Cielniak, Mark Swainson, and Dr Marc Hanheide. Courtesy of Chris Vaughan/ University of Lincoln.

Robotics specialists in the U.K., Sweden, Italy and Germany are working together to bring next generation automated guided vehicles (AGVs) into current warehouses to support tasks such as packing, palletizing and transporting goods.

The four-year project, called ILIAD — Intra-Logistics with Integrated Automatic Deployment — will deliver significant technological advances into a single integrated system ready for easy, low-cost deployment and without the need for major infrastructure investments.

Each robot will be human-aware — equipped with advanced computer vision and artificial intelligence to detect, track and predict human behavior. They will also be able to plan movements based on the machines' own observations of warehouse lay-outs and patterns of activity.

The robotic vehicles will also be self-optimizing, learning from self-collected data over time, making the fleets fully scalable with the option of adding or removing robots at any time.

While the aim is to develop a reliable robotic solution flexible enough to support intra-logistics in many different industries, the researchers will use the fresh food sector as their development setting. This is due to the industry's particularly challenging requirements: short shelf life of products, need for complete traceability, high cost of wastage, and pressure for rapid responses to changing market needs.

The laser technology that helps the forklifts operate autonomously is demonstrated on the screen, which is picking up nearby objects and translating it into a computerized diagram. Courtesy of Chris Vaughan/ University of Lincoln.

A team from the University of Lincoln is focused on long-term operation of the ILIAD system, including maintenance of environment maps over time and learning and predicting activity patterns of human co-workers. The team will also develop qualitative models for human-robot spatial interaction, systems architecture and systems integration.

"The fleets will be self-deploying and self-optimizing, removing much of the capital cost and disruption of introducing robotic technologies,” said Tom Duckett, director of the Lincoln Centre for Autonomous Systems at the University of Lincoln. “Most importantly, though, we will show that autonomous vehicles can operate safely and efficiently alongside human co-workers and human driven vehicles in complex, dynamic warehouse settings."

ILIAD is funded with a 7-million-Euros grant from the EU's Horizon 2020 fund. The consortium is led by Örebro University in Sweden, and includes the University of Lincoln, the University of Pisa, Italy, and Leibniz University in Hannover, Germany. The universities are working with major industrial partners Bosch, Kollmorgen Automation, ACT Operations Research, Logistic Engineering Services and Orkla Foods.