Abstract
Smart cities, societies, and organizations rely on automation of ordinary and advanced tasks to drive higher efficiencies and economic growth. Automatic personal identification is one such important task in modern enterprises such as for attendance systems and fingerprint verification is among the most reliable methods for the purpose. Designing a fingerprint-based attendance system has several challenges including the need to address false and multiple fingerprint logging, lack of necessary integration with other systems, etc. In this paper, we present a smart attendance system that uses fuzzy logic and integrates with the human resource system including a vacation database to address related challenges. The attendance system is implemented using .NET C# Web service, Oracle database, and the fingerprint device, ZKTeco U260-C. The implemented system has been deployed in a real environment in an organization comprising over 18,000 employees and has been tested using real data containing 6.1 million records. The proposed system is able to reduce the daily error rate from 14.7% to 2.5%. Conclusions are drawn with directions for future work.