Mitsubishi FUSO (MFTBC) is committed to developing a Factory of the Future. Plans are underway to improve the plant structure, process efficiency, and digitalization by 2025.

Along with various internal efforts, an important pillar of this strategy involves co-creation with bright and innovative students.

As an example of this, in 2021, the MFTBC Logistics Automation and Digitalization Team, together with a multi-national group of master’s students from ESADE University in Spain, jointly developed an AI algorithm for fully automated mobile robots.

A key initiative over recent years at FUSO has been to foster collaborations with academic organizations and startups for innovation. With ESADE, this matched the university’s desire to create opportunities for students to gain hands-on experience with industry experts. FUSO’s project allowed them to apply their data analysis and coding skills to a real-life industry challenge related to automated production.

Project background

MFTBC was looking to automate some of its operational processes in its truck manufacturing facilities. The purpose was to reduce the burden on its plant workers and help mitigate labor shortages.

One such case was the process of replenishing parts in the assembly line, which was previously not fully automated. 

While robots were involved in a portion of the process, they could only transport parts to and from different locations. However, they could not make decisions on the desired quantities. Operators still needed to make judgements on the number of parts needed and then manually loaded them onto the robots.

Couldn’t robots be made to do both tasks? The search was on for automated solutions.

Collaborating across continents

Known as the Capstone Project, work began in April 2021.

The teams conducted weekly online meetings between Spain and Japan to identify the most promising collaboration topic. Through their discussions, they selected the inventory check, where they could implement a visual detection algorithm with machine learning for robot applications.

This would allow the existing transportation robots equipped with cameras to identify empty slots on shelves. The empty slots would then trigger the robots to replenish the parts, thereby eliminating the need for manual loading.

The MCTBC team provided IT support, data, and photos. They also sent a specialized camera to Spain for the students to work with. The teams carried out testing in FUSO’s Botlab and then took the robots to the factory floor for testing in June.

The teams identified accuracy as the most important metric to judge the success of the project. By their July deadline, the AI’s accuracy reached 99.5%.

The project demonstrated the feasibility of automating inventory checks within the robot supply operation.

The collaboration produced a more accurate solution than that proposed by external developers, at a fraction of the cost, while growing the subject matter knowledge for both MFTBC workers & ESADE students.

The project demonstrates the possible win-win outcomes stemming from Corporate-University co-creation and is a model for FUSO GreenLab projects.

MFTBC’s doors are open to universities and student groups

With the success of the collaboration with students from ESADE University, MFTBC is again seeking to work with students in all areas.

For further automation and digitalization in the field of manufacturing supply chain, FUSO’s logistics experts are seeking collaborators for the following projects:

1. Development of an easy-to-use app that allows drivers to utilize a vehicle’s GPS and camera scanning functions. This would create a link between each truck and its delivery contents to increase supply chain transparency.
2. Design and develop a simple platform for a secure exchange of logistics data with FUSO parts suppliers for improving supply chain management.
3. Develop a machine-learning algorithm to analyze and predict the estimated arrival time of delivery trucks containing parts destined for assembly. This will enable a more effective supply chain and contingency planning.
4. Develop a parts-tracking business intelligence platform with a Geographic Google Maps (or other GIS) API. A user-friendly UX is also needed for logistics planners and intuitive KPI visualization for management. This would help make the logistics process more efficient.

In parallel, Fuso GreenLab is working on developing its network of domestic and international university collaborations for a variety of projects. Through these collaborations, FUSO experts mentor students throughout the process. This includes providing advice, tools snd facilities, knowledge sharing, future internships, letters of recommendation, and even the opportunity to apply to work for Daimler Truck.

If you are interested in collaborating with MFTBC in any of these areas, please contact us.

Our door is always open.