Arockia Selvakumar Arockia Doss

Dr. Arockia Selvakumar Arockia Doss

Vellore Institute of Technology Chennai, Chennai, India
orcid

Research Focus & Metrics

Academic Career Length

18Years (2009-Present)

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

Representative Publications

View All Publications
30 May 2023 in IEEE Access
Verified
Journal Article
Open access
GSG. Seeja
G. Seeja
Arockia Selvakumar Arockia DossArockia Selvakumar Arockia Doss
Arockia Selvakumar Arockia Doss
BVBerlin Hency V
Berlin Hency V
Search and Rescue operations for victim identification in an unstructured collapsed building are high-risk and time-consuming. The possibility of saving a victim is high only during the first 48 hours, and then the prospect tends to zero. The faster the response and identification, the sooner the victim can be taken to medical assistance. Combining mobile robots with practical Artificial Intelligence (AI) driven Human Victim Detection (HVD) systems managed by professional teams can considerably reduce this problem. In this paper, we have developed a Transfer Learning-based Deep Learning approach to identify human victims under collapsed building environments by integrating machine learning classification algorithms. A custom-made human victim dataset was created with five class labels: head, hand, leg, upper body, and without the body. First, we extracted the class-wise features of the dataset using fine-tuning-based transfer learning on ResNet-50 deep learning model. The learned features of the model were then extracted, and then a feature selection was performed using J48 to study the impact of feature reduction in classification. Several decision tree algorithms, including decision stump, hoeffiding tree, J48, Linear Model Tree (LMT), Random Forest, Random Tree, Representative (REP) Tree, J48 graft, and other famous algorithms like LibSVM, Logistic regression, Multilayer perceptron, BayesNet, Naive Bayes are then used to perform the classification. The classification accuracy of the abovementioned algorithms is compared to recommend the optimal approach for real-time use. The random tree approach outperformed all other tree-based algorithms with a maximum classification accuracy of 99.53% and a computation time of 0.01 seconds.

17 October 2022 in IEEE Access
Verified
Journal Article
Open access
GSG. Seeja
G. Seeja
Arockia Selvakumar Arockia DossArockia Selvakumar Arockia Doss
Arockia Selvakumar Arockia Doss
BVBerlin Hency V
Berlin Hency V
Snake robots have been a topic of discussion among researchers for decades. They are potentially strong enough to bring substantial contributions to the fields which are unsafe/ narrow/ dirty/ hard reachable to human operators, such as inspections, rescue missions, firefighting, etc. Though the inventions of the wheel and legged mechanisms are amazing, they often fail when coming to these scenarios. Terrain adaptability is the vital essence of locomotion over constrained surfaces in biological snakes. But how this natural adaptability is accomplished in snake-like robots? Therefore, this paper focuses a study on factors behind the recreation of a physical snake, like the kinematics and dynamics modelling, mechanical design, and locomotion control approaches from existing literature. With their feature comparison, the simulators available for verifying the mathematical model and the feasibility of the mechanical design are also made for researchers new to the field.

1 January 2019 in FME Transactions
Verified
Journal Article
Open access
PDF file
PLPavan Kalyan Lingampally
Pavan Kalyan Lingampally
Arockia Selvakumar Arockia DossArockia Selvakumar Arockia Doss
Arockia Selvakumar Arockia Doss
Current treatment for patients suffering from cervical spine injuries is by static braces which causes discomfort and immobilization of the neck. This paper presents a novel comfortable wearable therapy device for treating patients suffering with head/neck posture problems using parallel mechanism concept. To achieve this, a conceptual design of a three-degrees-of-freedom Revolute-Prismatic-Spherical configuration parallel manipulator with 3 extensible links is proposed. However, precision and workspace are the general considerations for designing a parallel manipulator. In this research, workspace and position analysis is carried out for the device. By applying geometrical method and vector-loop approach, the inverse kinematics equations are derived. The parallel manipulator is modeled using SolidWorks and simulated using MATLAB. From the simulation results, stroke length of the actuators, angular tilt and position of the moving platform is obtained. To validate the simulation results a preliminary prototype is built and tests were carried. The obtained results found to be satisfying the cause of the research. Furthermore, the device needs to automated and safely operated under the therapist supervision.

1 January 2018 in Procedia Computer Science
Verified
Journal Article
Open access
JFJanet J Fernandes
Janet J Fernandes
Arockia Selvakumar Arockia DossArockia Selvakumar Arockia Doss
Arockia Selvakumar Arockia Doss

20 July 2018 in Procedia Computer Science
Verified
Journal Article
Open access
PDF file
MMM. Manjunatha
M. Manjunatha
Arockia Selvakumar Arockia DossArockia Selvakumar Arockia Doss
Arockia Selvakumar Arockia Doss
VGVivek P Godeswar
Vivek P Godeswar
Manimaran RenganathanManimaran Renganathan
Manimaran Renganathan
Underwater robots are used in surveillance, academic and military applications because of their capability to execute risky tasks underwater without human intervention. In this study, a robot is built at low cost budget to carry out inspection at specified depth in a water body. Preliminary studies are performed to understand the buoyancy and stability of an underwater robot. Different configurations of propeller arrangements are identified and the final prototype is arrived after meeting the specific application requirements. Dual grippers are incorporated with the underwater robot, one for holding the pipe to avoid underwater currents and other for gripping/releasing the pipe. Gyroscope sensor is embodied in the underwater robot to maintain stability from its feedback signals. A high definition camera is mounted in front of the underwater robot for live video and image capturing purposes. Thus the underwater robot is able to meet all its requirements with minimal frontal drag. For validation purpose, computational fluid dynamics (CFD) studies are carried out with dynamic meshes and boundary conditions. The deviation between CFD and experimental results is found to be within the acceptable limits. The detailed cost analysis for the construction and testing of this underwater robot is also briefly tabulated.

10 November 2022
Verified
Book Chapter
Closed access
PDF file
RERajendran Sugin Elankavi
Rajendran Sugin Elankavi
DDDinakaran D
Dinakaran D
Kuppan Chetty RAMANATHANKuppan Chetty RAMANATHAN
Kuppan Chetty RAMANATHAN
MRM. M. Ramya
M. M. Ramya
Arockia Selvakumar Arockia DossArockia Selvakumar Arockia Doss
Arockia Selvakumar Arockia Doss