A six-month project to test the use of AI robotics in mail sorting has been completed at PostNord’s Rosersberg terminal in Sweden, offering promising results for the future of the post’s operations.
The pilot project was designed by the Nordic Technology Committee (NTC) and used an industrial robotic arm to sort bulk mail and varubrev (next-day) mail.
Carolin Schaffert, technology lead, automation, for network and technology strategy at PostNord Group, who managed the project, commented, “The AI robot pilot project is an important step in our efforts to meet the automation needs of today and to be well prepared for the needs of tomorrow.”
Intelligent sorting
Using sophisticated camera technology, the AI robot identifies and sorts small parcels after a cage turner. The camera takes a picture of the bulk flow, the AI network then identifies which parcel the robotic arm should pick first, and the arm then places this parcel in the right place on a conveyor belt. The AI robot also checks whether the parcel is ‘machine friendly’ or it should proceed to manual handling.
Magnus Lundin, head of machinery and maintenance at PostNord Sweden and a member of the NTC, explained, “We have been working with low-level automation for a number of years, but robotics technology is now ready to take this to the next level. I think automation and AI are key for enabling us to be profitable and an attractive option in the market in the long term.
“The AI robot is extremely efficient, roughly twice as efficient as a human, and if you manage the system smartly the robot can work 24/7 with a very low margin of error. This contributes to increased productivity while reducing production costs in a long-term perspective.”
Improved workplace
The pilot project also explored how the AI robot could be used to enhance the work environment and improve employee safety.
“Automation of repetitive tasks is an important issue for us. Monotonous tasks and poor ergonomics can cause repetitive strain injuries. With smart technology solutions such as this, these kinds of problems can be avoided,” continued Schaffert.
Hampus Schander, PostNord’s sorting blueprint workstream lead, added, “Robotic singulation makes it possible to feed parcels into an existing sorting facility in a cost-effective way and in a limited space. This reduces costs, which is one of the main aims of our Cost Leadership program. The fact that the robot reduces the amount of manual lifting also helps improve our work environment.”
Future of AI
The outcome of the AI robot pilot project has been described as a success, with the final parts of the pilot project now being evaluated.
“The team is analyzing how to increase the number of robots in mail terminals in a next step. In addition, they are also looking at the possibilities of using a similar AI robot for the parcel network, which is of interest for all countries. We have lots of exciting AI projects underway. There is huge potential for us in the future with regard to AI and automation,” Schaffert concluded.