Ajay Atmaram Waghmare

Mr. Ajay Atmaram Waghmare

Electrician and Computer Engineering, Oakland University, Rochester, Indiana, USA
orcid
Oakland university

Research Focus & Metrics

Academic Career Length

3Years (2024-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
14 August 2024 in Preprints.org
Verified
Preprint
Open access
PDF file
Ajay Atmaram WaghmareAjay Atmaram Waghmare
Ajay Atmaram Waghmare
Subramaniam GanesanSubramaniam Ganesan
Subramaniam Ganesan
JCJingshu Chen
Jingshu Chen
The future of transportation is being reshaped by autonomous vehicles (AVs), which promise to revolutionize road safety, accessibility, and environmental sustainability. Central to their evolution are advancements in artificial intelligence (AI), which enable real-time perception, decision-making, and control. This paper explores the future trajectory of AI in AVs, highlighting innovative AI paradigms such as explainable AI, federated learning, and adaptive regulatory frameworks. It delves into the development of sophisticated perception algorithms for environmental understanding, advanced decision-making systems employing reinforcement learning and deep neural networks, and precise control mechanisms through predictive modeling and optimization. By addressing current challenges, including regulatory hurdles and the complexity of AI algorithms, this paper provides a comprehensive outlook on the potential and implications of AV technology. It envisions a future where AI-driven AVs are seamlessly integrated into society, offering profound insights into their impact on transportation and societal dynamics.
1

17 July 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
Ajay Atmaram WaghmareAjay Atmaram Waghmare
Ajay Atmaram Waghmare
NANoor Hasan Ahmed
Noor Hasan Ahmed
Subramaniam GanesanSubramaniam Ganesan
Subramaniam Ganesan
Advanced Driver Assistance Systems (ADAS) rely on tightly integrated sensing, estimation, and control pipelines to enhance road safety while maintaining the human driver in the control loop. This paper presents a unified, reproducible modeling and evaluation framework for four widely deployed ADAS functions—Lane Keeping Assist (LKA), Forward Collision Warning (FCW), Blind Spot Detection (BSD), and Automated Parking Assist. For each function, we detail the sensing architecture, governing mathematical models, decision and control logic, and representative validation criteria drawn from industrial practice. The models include a kinematic bicycle formulation for lateral control, constant‑velocity and constant‑acceleration time‑to‑collision kinematics for collision‑risk estimation, polygonal zone geometry for radar‑based blind‑spot monitoring, and a two‑arc geometric construction for parking path planning. Reference Python implementations and simulation results are provided, demonstrating RMS lateral offset of 0.29 m for LKA, FCW alert latency of 50 ms, BSD zone dwell accuracy within 0.5 s, and automated parking maneuver time of 4.83 s. A Monte‑Carlo sensitivity analysis of FCW thresholds illustrates the trade‑off between early warning and nuisance‑alert rate that governs production calibration. The unified treatment offered here provides a consolidated reference for researchers and engineers developing or evaluating ADAS sensing‑to‑actuation pipelines.
1

11 August 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
Ajay Atmaram WaghmareAjay Atmaram Waghmare
Ajay Atmaram Waghmare
Subramaniam GanesanSubramaniam Ganesan
Subramaniam Ganesan
Modern vehicles carry front and rear radar, forward and surround-view cameras, and increasingly infrared sensing as standard driver-assistance equipment. This paper investigates whether that same sensor suite, fused and fed forward into the suspension control loop, can measurably improve ride quality on unpaved and off-road surfaces, for both the vehicle chassis and, as a secondary actuation layer, the occupant seat. We formalize a sensor-fusion, road/suspension estimation, control, and feedback architecture into a state-space model; derive the governing equations for a three-degree-of-freedom quarter-car-plus-seat plant; specify a multi-sensor preview measurement model and an inverse-variance fusion law; design and compare passive, reactive, and predictive controllers; and implement and evaluate the full pipeline in Python on a synthetic ISO 8608 [2] off-road profile with embedded speed-bump test events. The predictive controller reduces ISO 2631-weighted [1] body (chassis) RMS acceleration by 23.7% and seat/occupant RMS acceleration by 67.9% relative to a passive baseline. We further replace two hand-engineered blocks with learned counterparts — a neural network that corrects context-dependent sensor bias, and a Random Forest terrain classifier that drives gain scheduling — and report the resulting, honestly modest, additional gains. A prior-art survey spanning patents [4–7,10,11] and production systems [8,9], and a review of public datasets [12–14,18] suitable for validating this class of system, precede an explicit novelty and freedom-to-operate discussion.
1

29 September 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
Ajay Atmaram WaghmareAjay Atmaram Waghmare
Ajay Atmaram Waghmare
Diabetic retinopathy (DR) is a leading cause of preventable blindness, yet large numbers of people with diabetes miss their annual eye examination because grading retinal photographs requires scarce specialists. This paper proposes an artificial intelligence (AI) based eye-inspection framework that embeds automated DR screening inside a routine general medical checkup. The framework comprises five layers: image acquisition, automated quality control, lesion feature extraction, a stacked machine learning (ML) decision engine, and an explainable clinical report with referral. We evaluate the approach on the public UCI Diabetic Retinopathy Debrecen database, which contains 19 lesion, anatomical and image-level features extracted from 1,151 MESSIDOR fundus images. Nine classical and neural classifiers are benchmarked with 10-fold stratified cross-validation, after which the proposed pipeline adds lesion-aware feature engineering, hyperparameter tuning and stacked generalization. The proposed model reaches a cross-validated area under the ROC curve (AUC) of 0.854 and accuracy of 77.6%, improving the logistic-regression baseline by 0.029 AUC, and achieves a hold-out AUC of 0.878 (95% bootstrap CI 0.830–0.917) with 79.7% sensitivity and 77.6% specificity. Lowering the decision threshold to 0.30 yields 91.9% sensitivity for screening use. In an extension to cataract, three image-clarity features separate normal from cataract fundus images with a cross-validated AUC of 0.978. Permutation importance and ablation show that microaneurysm counts carry most of the diagnostic signal, consistent with clinical grading. Using CDC surveillance data covering 9.6 million Americans with DR, we project that the screening operating point would detect 8.8 million cases while removing about 16 million specialist grading visits per year. We discuss clinical workflow integration, challenges including domain shift, bias and regulation, and a roadmap towards deep-learning-based multi-disease retinal screening.

22 April 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
Ajay Atmaram WaghmareAjay Atmaram Waghmare
Ajay Atmaram Waghmare
Subramaniam GanesanSubramaniam Ganesan
Subramaniam Ganesan
Automotive electronic control units increasingly rely on resource-constrained microcontrollers to implement authentication and message validation for in-vehicle networks such as CAN. Although cryptographic primitives may be mathematically secure, their software implementations on low-cost ECUs can leak sensitive information through timing side-channels. This paper presents a practical timing side-channel evaluation of an embedded authentication routine implemented on Arduino-class microcontrollers representative of entry-level automotive ECUs. A Python-based measurement framework interacts with the target device over a serial interface, repeatedly triggers authentication operations, and records execution time under controlled conditions. Experimental results show statistically distinguishable timing distributions correlated with secret-dependent execution paths, enabling an attacker with CAN-adjacent access to infer partial information about authentication checks. Lightweight countermeasures, including constant-time comparisons and control-flow normalization, are implemented and evaluated. The mitigated design reduces observable timing leakage with minimal execution-time and memory overhead, highlighting the need for implementation-level timing analysis in automotive embedded systems.
1

3 August 2026 in Preprints.org
Verified
Preprint
Open access
PDF file
Ajay Atmaram WaghmareAjay Atmaram Waghmare
Ajay Atmaram Waghmare
NANoor Hasan Ahmed
Noor Hasan Ahmed
Subramaniam GanesanSubramaniam Ganesan
Subramaniam Ganesan
Accurate and real-time road sign recognition is a safety-critical capability in autonomous driving systems, where detection latency directly impacts vehicle response and passenger safety. This paper presents an AI-driven, fully Python-based Optical Character Recognition (OCR) pipeline for autonomous traffic sign detection, incorporating a probabilistic Region of Interest (ROI) scanning strategy guided by historical spatial detection maps and a Long Short-Term Memory (LSTM) temporal predictor. Rather than processing full camera frames at high computational cost, the proposed system intelligently narrows its scanning focus to frame zones with the highest statistical likelihood of containing traffic signs, reducing the effective pixel evaluation area by up to 78% per frame. The pipeline achieves a mean Average Precision (mAP) of 97.8% at an inference latency of 31 ms on CPU and 10.3 ms on GPU — a 78% latency reduction and 6.6 pp accuracy improvement over the full-frame scanning baseline. A continuous AI self-improvement module employing Elastic Weight Consolidation (EWC) and memory replay enables incremental retraining from 94.2% to 98.4% validation accuracy across six rounds with negligible catastrophic forgetting.
1