Predicting Time Pressure of Powered Two-Wheeler Riders for Proactive Safety Interventions PubID: pubid:

Sumit S. Shevtekar, Chandresh K. Maurya, Gourab Sil, Subasish Das · 2026

Read Predicting Time Pressure of Powered Two-Wheeler Riders for Proactive Safety Interventions PubID: pubid: by Sumit S. Shevtekar, Chandresh K. Maurya, Gourab Sil, Subasish Das (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2601-03173.txt.

Introduction

Sumit S. Shevtekar, Chandresh K. Maurya, and Gourab Sil build a model that predicts when motorcycle and scooter riders are riding under time pressure, a state that provokes risky maneuvers. Time pressure cannot be measured directly on the road, so the authors infer it from riding data. Their model, MTPS, combines convolutional preprocessing, two stages of temporal attention, and feature recalibration, yet stays tiny enough to run in a fraction of a millisecond.

To train and test it, they collected a dataset of more than one hundred twenty-nine thousand feature windows from simulator sessions with fifty-one experienced riders under no, low, and high time pressure. High pressure made riders ride much faster, brake more suddenly, and take far more risky turns. The predictor beat six baselines and is demonstrated for collision prediction.

Catalog record

CREATORSumit S. Shevtekar, Chandresh K. Maurya, Gourab Sil, Subasish Das
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE59 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2601-03173

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Predicting Time Pressure of Powered Two-Wheeler Riders for Proactive Safety Interventions PubID: pubid: Sumit S. Shevtekar, Chandresh K. Maurya, Gourab Sil, Subasish Das Abstract Time pressure critically influences risky maneuvers and crash proneness among powered two-wheeler riders, yet its prediction remains underexplored in intelligent transportation systems. We present a large-scale dataset of 129,000+ labeled multivariate time-series sequences from 153 rides by 51 participants under No, Low, and High Time Pressure conditions. Each sequence captures 63 features spanning vehicle kinematics, control inputs, behavioral violations, and environmental context. Our empirical analysis shows High Time Pressure induces 48% higher speeds, 36.4% greater speed variability, 58% more risky turns at intersections, 36% more sudden braking, and 50% higher rear brake forces versus No Time Pressure. To benchmark this dataset, we propose MotoTimePressure, a deep learning model combining convolutional preprocessing, dual-stage temporal attention, and Squeeze-and-Excitation feature recalibration, achieving 91.53% accuracy and 98.93% ROC AUC, outperforming eight baselines. Since time pressure cannot be directly measured in real time, we demonstrate its utility in collision prediction and threshold determination. Using MTPS-predicted time pressure as features, improves Informer-based collision risk accuracy from 91.25% to 93.51%, approaching oracle performance (93.72%). Thresholded time pressure states capture rider cognitive stress and enable proactive ITS interventions, including adaptive alerts, haptic feedback, V2I signaling, and speed guidance, supporting safer two-wheeler mobility under the Safe System Approach. I Introduction Powered two-wheelers (PTWs) are among the most accessible and widely used transport modes, particularly in developing countries. However, they carry elevated crash risks due to exposure to rider and balance control demands [ 8 ] . The severity of the crash is strongly influenced by overspeeding, traffic dynamics, rider behavior, and road conditions [ 49 , 7 , 11 ] , with human error and risky maneuvers being the leading causes [ 38 , 4 , 14 ] . PTWs account for 21% of global road fatalities rising to 38% of deaths and 51% of injuries in Low- and Middle-Income Countries (LMICs) [ 55 ] . Almost half of traffic deaths involve vulnerable users such as pedestrians, cyclists, and PTW riders [ 55 ] . Excess speed alone contributes to 34% PTW fatalities, with riders traveling up to 2.3 times faster than car drivers [ 50 ] . According to India’s Ministry of Road Transport and Highways (MoRTH), PTWs are the most prevalent and economical mode of transport in LMICs. Between 2003 and 2022, India’s registered vehicle population grew to 354 million, of which PTWs constituted 263 million (74.4%). This segment grew at a CAGR of 8.6%, surpassing buses (2.48%) and goods vehicles (7.3%). Category-wise registration trends are shown in Fig. 1 , underscoring the growing dependence on PTWs and the urgent need for targeted safety interventions [ 36 ] . Fig. 1: Registered motor vehicles in India by category (2003–2022) [ 36 ] . PTWs also account for the largest share of traffic fatalities in India, accounting for 44.5% of deaths in 2022, followed by pedestrians (19.5%), cars/LMVs (12.5%), trucks/lorries (6.3%), autorickshaws (3.9%), bicycles (2.9%), buses (2.4%), and others (8.0%) [ 28 , 26 , 27 ] . According to official MoRTH reports on road accidents in india , road fatalities in India are predominantly male, ranging from 85.2% to 87.3% between 2019 and 2023 [ 31 , 32 , 33 , 34 , 35 ] , with most fatalities occurring in the 18–45 age group. These statistics show the high risk associated with male PTW riding behaviour. Driving errors are the main cause of crash globally, especially for PTWs, which face high instability and dynamic control challenges [ 5 , 9 ] . Field studies of PTW behavior have shown significant variability in acceleration and braking patterns [ 30 ] , emphasizing the inherent dynamic control challenges these vehicles face in real-world traffic. Simulator studies further reveal context-dependent rider behavior in mixed traffic [ 59 ] , underscoring PTWs vulnerability and the need for targeted safety interventions. A critical yet underexplored factor in PTW safety is time pressure (TP), reflecting perceived urgency or limited time to complete riding tasks [ 42 ] . For example, many delivery PTW riders in India are subjected to TPs [ 51 , 46 , 47 ] . Elevated TP impairs cognition and motor control, driving overspeeding, braking inconsistencies, delayed hazard response, and higher crash risk [ 42 , 17 , 50 ] . Naturalistic studies link risky behavior to speed, maneuvering, and context [ 21 ] , while stress further amplifies variability and control errors [ 38 ] . Field evidence from Thailand also confirms that widespread speeding significantly raises collision risk [ 49 ] . Pawar et al. [ 42 ] showed that TP induces risky maneuvers. However, directly measuring TP in real-world PTW riding is practically impossible. We hypothesize that TP can be predicted from PTW’s data. Therefore, we established a proxy measure for TP and predict it using other PTW telematic features such as speed, acceleration, etc. Towards this end, we first collect data in the same fashion as [ 42 ] but for PTW. That is, riders experienced TP conditions (no time pressure (NTP), low time pressure (LTP), high time pressure (HTP)) and asked to drive PTW. Thus, we collect the supervised data. To the best of our knowledge, this is the first work that develops predictive models for cognitive state detection in PTW riders, addressing a fundamental limitation in current ITS. To evaluate the data quality, we propose a novel MotoTimePressure (MTPS) model that predicts TP. To show the utility of such a predicted TP feature in collision risk prediction, we train an Informer [ 60 ] model using the ground truth (GT) TP feature and the predicted TP feature. Our results show that predicted TP features increase the accuracy of collision risk prediction from 91.25% to 93.51%, approaching oracle performance (93.72%). The key contributions of this work are summarized as follows: 1. The First Labeled Dataset for PTW Rider Cognitive State : To the best of our knowledge, We present a first high-resolution PTW simulator dataset comprising more than 129,000 labeled multivariate time-series sequences from 153 rides by 51 participants under three TP conditions (NTP, LTP, HTP). Each sequence includes 63 features covering vehicle kinematics, control inputs, behavior violations, and environmental context, providing a rich resource for modeling cognitive load and rider safety. 2. Speed and Control Degradation Under TP : We provide comprehensive statistical evidence that TP systematically degrades riding control. Under HTP conditions, riders exhibit 48% higher mean speeds and 36.4% greater speed variability compared to NTP. Control degradation extends to 604% increased front brake force application, 50% higher rear brake forces, and 35% increased gear usage, demonstrating systemic impairment of fundamental riding skills under TP. 3. Systemic Risk Escalation and Inadequate Compensation : We demonstrate that TP creates compound collision risks across multiple riding domains. HTP causes: 58% more risky turns at incorrect speeds, 36% more sudden braking incidents, 50% higher brake forces, and 67% worse clutch control. Despite natural compensation behaviors, human adaptation fails as HTP still causes 21.3% worse control variability and 10.2% more severe safety violations. Statistical analysis reveals a consistent degradation pattern: TP \rightarrow degraded control \rightarrow insufficient self-correction \rightarrow persistent collision risks. 4.

Read the complete text