Thesis Engineering Student : Inbound pallet inspection with AI on Movu controller

Problem
Our automated warehouse solutions use shuttles to move pallets in and out of our racks. The pallets must have a correct form factor and decent quality to ensure smooth handling. Also the load on the pallet has to meet predefined criteria to avoid incidents.
Hence, quality inspection of pallet and load is a must before the pallet enters the warehouse.

Main issues
• Wooden pallets are partly damaged and might not hold the load during their life cycle. Over time or during handling the pallet could break and the load could fall down.
• The pallet type has dimensions that are outside the tolerance of the shuttle specification. The pallet could get stuck while the shuttle is driving around or might not be supported sufficiently when placed in the rack.
• The wrapping foil around the load is hanging loose over the pallet. The optical detection sensors of the shuttle could detect this foil as a solid object blocking the way and as such preventing the pallet of being picked.
• The load is unequally placed on the pallet and hanging over.
The cases mentioned above affect the performance of the warehouse solution in multiple ways:
- Performance drop: downtime of a part of the warehouse
- Costs: damage to goods and racking structure
- Danger: unsafe working conditions for maintenance employees in the rack

Solution idea

Today Movu has a pallet inspection setup that uses camera’s with AI algorithms to detect unallowed pallets or load configurations. An NVIDIA Jetson powered processing platform is used in this setup. Recently Movu has designed its own controller platform (MOC) for use inside the shuttles. This thesis focuses on how the MOC can be used as replacement for the Jetson platform.

Required competences:

• Bachelor last year or Master Engineering / Software
• You live in Belgium and speak Dutch or English
• You work at least 4 days in the office in Lokeren
• Minimum duration of 3 months

What we offer you?

• A place in an enthusiastic, young organisation with the necessary ambitions
• The necessary freedom and the opportunity to take initiatives
• Location Office : Lokeren

Unsolicited representations by third parties (recruitment agencies, headhunters, ...) of CVs via mail and/or telephone for our vacancies are considered as direct applications where no compensation is provided to the third party. Any T&Cs from these third parties will not be accepted unless upon signature of the T&Cs by a person in charge of HR. Candidates remain registered in the system for 12 months and cannot be proposed again during this period.