EPIC recognized early that this project carried significant technical risk.
Traditional rule-based vision systems were not designed for this level of product variability, and inconsistent imaging conditions could dramatically impact inspection reliability.
EPIC initiated a proof-of-concept and Front-End Engineering (FEE) phase to:
- Validate the feasibility of real-time label identification
- Determine whether lighting consistency could be achieved across all label types
- Evaluate imaging performance under real-world conditions
- Define integration requirements with existing controls and sortation systems
- Reduce implementation risk before deployment
The goal was not simply to test components—it was to engineer a repeatable, scalable inspection strategy capable of supporting long-term production demands.
EPIC used its internal Vision Lab to simulate actual production conditions using real customer samples.
Over several months, the engineering team conducted iterative testing focused on:
- Lighting configurations
- Imaging consistency across reflective and low-contrast labels
- Camera resolution and field-of-view requirements
- Barcode readability under varying conditions
- Inspection repeatability across multiple stations
- Software architecture and centralized processing strategies
The team evaluated multiple vision software platforms and ultimately selected a PC-based architecture capable of supporting:
- Centralized decision-making
- Multiple synchronized inspection stations
- Advanced image analysis
- High-speed processing requirements of running 4000 SKUs concurrently
During testing, EPIC identified lighting consistency—not camera capability—as the primary driver of system reliability. This insight directly shaped the final system design.
Controlled Imaging Environment
To eliminate variability at the source, the team designed a fully enclosed inspection system:
- Removed interference from ambient plant lighting
- Created a repeatable and stable imaging environment
- Ensured consistent conditions regardless of external factors
Advanced Lighting Strategy
A large dome lighting configuration was implemented to address the wide range of label conditions:
- Delivered uniform, diffuse illumination across a large field of view
- Minimized harsh reflections and hotspots on glossy surfaces
- Balanced brightness to accommodate both low-contrast matte labels and reflective materials
Lighting angles, intensity, and diffusion were carefully engineered to reduce glare while maintaining sufficient contrast for accurate detection.
Standardization Across the System
Lighting design was standardized across all inspection stations, ensuring:
- Consistent image quality throughout the system
- Simplified maintenance and troubleshooting
- Repeatable performance across multiple facilities
Scope Definition & Integration Planning
The proof-of-concept phase extended beyond imaging validation. EPIC also conducted on-site evaluations to define project scope and integration boundaries.
The engineering team worked directly with plant personnel to:
- Review existing conveyors and sortation systems
- Identify controls integration points
- Determine which equipment would remain versus be replaced
- Document electrical and mechanical requirements
- Define communication between inspection and sortation systems
This process established a clear implementation roadmap while minimizing disruption to existing operations.
