onsite
Thesis project: Edge-AI on embedded MCU using specialized model
Embedded Systems Engineer
A master thesis project (2 students, 30 HP) at Toyota Material Handling in Mjölby: implement pre-trained and custom-trained tiny task-specific AI models on an STM32N6 embedded MCU for forklifts, integrating with existing systems to enable features like voice/gesture commands, pallet analysis, and obstacle detection.
About the role
The potential of tiny, locally ran models, with distinct tasks, is largely unexplored. Toyota Material Handling Europe seeks an investigation into the possibilities enabled by next-generation hardware, when applied to the context of material handling vehicles.
In your Master Thesis at Toyota, you will work on:
Implementation of pre-trained, task-specific models on a target MCU (stm32n6).
Tailoring and training of custom task-specific models
Integration with existing systems to enable new functionality
Collaboration with another thesis work regarding exploration of possible use-cases.
Implementing and integrating into manually driven forklifts, tiny task-specific models that may be used to enable new functionality. Examples of use-cases to explore include voice-command, gesture command, pallet analysis, obstacle detection and many more.
Scope
2 individuals, 30 HP.
For this master