Machine learning detects e-scooter rider intoxication via kinematic analysis

Researchers have developed a method to detect alcohol intoxication in e-scooter riders using sensor data and machine learning. By instrumenting an e-scooter with an Inertial Measurement Unit (IMU) and other sensors, they collected data on participants' riding patterns while sober and at different blood alcohol concentration levels. Analysis of kinematic features, specifically normalized permutation entropy and standard deviation of IMU and throttle signals, revealed that intoxication leads to a shift from continuous micro-corrections to fewer, larger reactive corrections. A logistic regression model trained on these features achieved 85% accuracy in distinguishing between sober and intoxicated riders, with steering rate and lateral acceleration identified as key predictive indicators. AI

IMPACT Could enable real-time safety systems for personal mobility devices, reducing accidents caused by impaired operation.

RANK_REASON Research paper detailing a novel application of machine learning for a specific detection task. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• alcohol intoxication
• arXiv
• logistic regression model
• machine learning
• Rahul Rajendra Pai

• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →