- The Factory Copilot project was conceived to revolutionize industrial assembly supervision by integrating AI-enabled wearable technology into the workflow of automotive manufacturing. The goal was to automate quality control for tasks such as wiring and nut fastening, improving both efficiency and accuracy. The project underwent three major iterations, each enhancing the design, integration, and performance of the device.
- The hardware design featured an ergonomic wearable unit capable of housing the computing and sensing modules. At the core of the system was a Raspberry Pi Zero, chosen for its compact size and sufficient computational power for embedded AI inference. An Arducam camera module was integrated to capture real-time visual data from the assembly line. This hardware combination enabled the deployment of lightweight object detection algorithms directly on the wearable device, allowing it to recognize assembly tasks and verify correctness without reliance on external systems.
- The AI component leveraged object detection models trained to identify key assembly actions and detect errors in real time. These models were optimized for embedded deployment, ensuring low latency and robust performance under factory conditions. The wearable device was designed to operate in dynamic, high-vibration industrial environments while maintaining ergonomic comfort for the operator.
- Finite Element Analysis (FEA) was conducted across all three iterations to evaluate mechanical durability and fracture resistance of stretchable components subjected to repetitive deformation. Iterative prototyping refined both mechanical and electrical subsystems, resulting in a compact and resilient form factor capable of housing all components securely without hindering operator mobility.
- Iteration 3 delivered a version with significantly improved ergonomics, comfort, and durability, incorporating extensive prototyping and user feedback. Its superior wearable design and operator usability were key factors that led to securing a contract with Haval for industrial deployment.
- The final deployed prototype significantly improved assembly operations, achieving a 50% reduction in cycle time and improving task supervision accuracy. This was accomplished through seamless integration of hardware, AI, and software, coupled with rigorous testing and refinement. The project required multidisciplinary collaboration, encompassing mechanical design, embedded systems engineering, AI model deployment, and manufacturing process integration, resulting in a robust, industrial-grade solution ready for real-world deployment.
HOME
Factory Copilot – Industrial Assembly Device (Iteration 3)
Project Timeline
Jun 2025 - Aug-2025
OVERVIEW
Factory Copilot is an AI-enabled wearable device developed to automate supervision in automotive assembly processes, specifically wiring and nut fastening. The system integrates custom hardware design, sensor fusion, and AI-driven recognition models to reduce assembly cycle time and improve process reliability.
My role focused on end-to-end hardware engineering, including mechanical design, finite element analysis (FEA) for stretchable components, and system integration of sensors with the wearable form factor. The project resulted in a deployment-ready prototype that demonstrated significant reductions in assembly time on the factory floor.
HighlightS
- Led complete hardware design of the wearable device, from concept to deployment-ready prototype.
- Conducted FEA to assess durability and fracture resistance of stretchable components under repetitive deformation.
- Integrated sensors and AI models into a compact, ergonomic form factor for industrial use.
- Reduced automotive assembly cycle time by 50%, enhancing efficiency in wiring and nut fastening operations.
- Collaborated with cross-functional teams across software and manufacturing to achieve seamless system integration.
- Delivered Iteration 3 with superior ergonomics, comfort, and durability, securing a Haval contract due to its optimized wearable design and enhanced operator usability.
SKILLS
Mechanical Design & CADFinite Element Analysis (FEA)Embedded Hardware IntegrationRapid PrototypingCross-Functional Collaboration
Additional Details
lowinertia
Portfolio Builder for Engineers