OTGP: An innovative biometric authentication system with on-the-go passwords using a novel typing signature

Traditional password systems are vulnerable to cyberattacks such as brute force and phishing attacks, while existing biometric methods like fingerprint and facial recognition require specialized hardware and can be compromised. Keystroke dynamics offers a promising alternative by analyzing unique typing patterns using standard keyboards without additional equipment, leveraging motor memory that is difficult for attackers to replicate. We hypothesized that typing patterns are unique to every individual and can serve as a reliable biometric authentication method. To test this, we developed specialized hardware using piezoelectric sensors integrated into a conventional keyboard, along with custom software that captures five typing parameters: typing speed, inter-key intervals, typing accuracy, keystroke duration, and keypress force. These parameters generate what we term a "typing signature" for each user. Our method enables On-The-Go Passwords—phrases dynamically created by users at their discretion. We conducted an experimental study with 30 participants from various demographic backgrounds across six sessions each to evaluate consistency, uniqueness, and reliability. Statistical analysis confirmed our hypothesis: typing signatures showed high consistency within individuals (mean Coefficient of Variation < 15%), clear distinctiveness between individuals (99.8% of pairwise comparisons showed significant separation), and significant age correlations for temporal and force metrics (p < 0.05), while maintaining discriminative power across demographic groups. These findings suggest typing signatures can serve as an effective biometric authentication method with practical applications in finance and healthcare.