Lakshman Mahto

About

My research centers on advancing learning-driven autonomy in decision systems that must perform reliably amid dynamic and uncertain environments. I develop end-to-end architectures that integrate perception, prediction, and planning, grounded in dynamical systems theory, statistical learning, and optimization. My group builds multimodal generative predictors (e.g., conditional variational autoencoders), risk-aware planners (differentiable model predictive control with certified safety and stability guarantees), and scalable solvers (convexification techniques and distributionally robust optimization). These tools enable applications in intelligent mobility-including autonomous and assisted driving, traffic flow management, and logistics optimization as well as socio-economic dynamics, such as migration and stress forecasting, crime risk modeling, and equitable resource allocation. Our emphasis is on safety, robustness, and interpretability, achieved through formal guarantees and reproducible, open-source pipelines, with a clear trajectory toward real-world deployment in collaboration with institutional partners.

Employment

• Birla Institute of Technology, Mesra Assistant Professor
  2024 - Present
• Indian Institute of Information Technology Dharwad Assistant Professor
  2016 - 2024
• Institute of Mathematical Sciences Postdoctoral Fellow
  2015 - 2016
• Indian Institute of Technology Mandi Project Assistant
  2014 - 2015
• Birla Institute of Technology Senior Research Fellow
  2010 - 2010

Education

• Indian Institute of Technology Mandi PhD
  2010 - 2014
• Ranchi University M.Sc.
  2007 - 2009
• Saint Xaviers College B.Sc.
  2004 - 2007
• Saint Xaviers College B.Sc.
  2004 - 2007
• Saint Xaviers College B.Sc.
  2004 - 2007

Projects & Funding

Projects & funding information is unavailable.

Publications (15)

• Conditional Variational Autoencoder Networks for Autonomous Vehicle Path Prediction
  Neural Processing Letters 2022 DOI: 10.1007/s11063-022-10802-z
  +Save
• Deep Learning Techniques for Autonomous Vehicle Path Prediction
  AAAI workshop on the AI for Urban Mobility Workshop 2021
  +Save
• Detection of Reckless Driving using Deep Learning
  2020
  +Save
• An approximate gradient-based hyper-parameter optimization in a neural network architecture
  2020
  +Save
• Density Based Clustering Methods for Road Traffic Estimation
  2020 IEEE REGION 10 CONFERENCE (TENCON) 2020 DOI: 10.1109/tencon50793.2020.9293790
  +Save
• Approximate Controllability of Sub-Diffusion Equation with Impulsive Condition
  Mathematics 2019 DOI: 10.3390/math7020190
  +Save
• Piecewise Continuous Stepanov-Like Almost Automorphic Functions with Applications to Impulsive Systems
  Current Trends in Mathematical Analysis and Its Interdisciplinary Applications 2019 DOI: 10.1007/978-3-030-15242-0_4
  +Save
• Dynamical Study of Fractional Model of Allelopathic Stimulatory Phytoplankton Species
  Differential Equations and Dynamical Systems 2014 DOI: 10.1007/s12591-014-0219-5
  +Save
• Asymptotic almost automorphic solutions of impulsive neural network with almost automorphic coefficients
  2014 DOI: 10.1016/j.neucom.2014.04.028
  +Save
• PC-Almost Automorphic Solution of Impulsive Fractional Differential Equations
  2014 DOI: 10.1007/s00009-014-0449-3
  +Save