Enhancing labour efficiency in coffee harvesting using process mining and statistical analysis

Area of Science:

• Agricultural Science
• Industrial Engineering
• Data Science

Background:

• Coffee processing quality and market value are significantly impacted by operational challenges in wet processing environments.
• This study examines coffee harvesting and processing in Sakaleshpur (Arabica) and Chikkamagaluru (Robusta).
• The research includes 15 Robusta batches and three Arabica harvesting stages: fly picking, mid-harvesting, and stripping.

Purpose of the Study:

• To analyze process features and productivity-impacting factors in coffee harvesting.
• To investigate the effectiveness of integrating statistical analysis and Process Mining (PM).
• To provide a data-driven approach for enhancing coffee processing efficiency.

Main Methods:

• Utilized statistical analysis to examine process features and productivity factors.
• Employed Process Mining (PM) to identify workflow patterns and operational variations.
• Analyzed 15 batches of Robusta and three stages of Arabica harvesting.

Main Results:

• A negative correlation was found between the number of workers and coffee quantity during fly picking, indicating diminishing returns with increased labor.
• Statistical analysis identified labor efficiency, weather, and gender as key factors influencing harvesting performance.
• Process Mining (PM) effectively identified variations in workflow patterns, complementing traditional cause-and-effect research.

Conclusions:

• Integrating Process Mining (PM) and statistical techniques offers valuable insights into coffee harvesting processes.
• Effective labor management, alongside human and environmental considerations, is crucial for increasing productivity.
• This integrated methodology establishes a data-driven paradigm for optimizing coffee processing systems.