Postdoctoral Researcher — HUMAN Lab
Postdoctoral Researcher HUMAN Lab · University of California, Los Angeles
New Patent published — Personalised dynamic route planning system for sustainable urban mobility, 31 July 2026I am a transportation researcher. My work is about how people decide to travel — which mode, which route, at which hour — and how those decisions meet the environment they move through. What holds my attention is the moment behaviour changes: under air pollution, under heat, under a new piece of infrastructure or a new piece of information. Understanding when a traveller will shift, and who is constrained from shifting at all, is the thread running through everything I do.
Research interests
Method is where I spend most of my time. I design stated-preference and panel experiments, estimate hybrid latent class, ICLV and mixed logit models, and develop new estimation approaches — most recently PD‑MUSE. I also build the software the work runs on: survey instruments, routing engines and dashboards. The apparatus belongs to behavioural economics and decision analytics as much as to transportation engineering.
Ph.D. IIT Kharagpur (Prof. Arkopal Kishore Goswami) · M.Tech IIT Roorkee (Prof. Amit Agarwal) · previously WRI India and a research exchange fellow at the University of Leeds.
Methods
Primal–Dual Maximum-Utility Structured Estimator
Nested and inverse product-differentiation logit models buy realistic substitution patterns at the cost of a likelihood that is no longer concave. PD‑MUSE recasts estimation as a convex program, and confines what remains to a bounded, low-dimensional search.
Under review · Transportation Research Part B: Methodological
Fieldwork
Heat · Exposure · Activity · Travel
As cities warm, people shift trips earlier, shorten them, or change mode — and those who cannot absorb the heat instead. HEAT connects thermal stress to how days are scheduled and how journeys are actually made.
In progress · HUMAN Lab, University of California, Los Angeles
Direction
Learned models, held to a behavioural standard
Travel data now arrives as text, imagery, sensor streams and trajectories at once, and learned models can read all of it. I work on where that genuinely helps — predicting choices, filling sparse networks, reading the street from imagery — and on keeping the answer interpretable enough for someone to act on.
Ongoing · alongside the behavioural modelling
My work spans transportation engineering, environmental health, behavioural science and applied machine learning. Below: the programme, then what it has found.
One question, approached from four directions: how do people decide to move, and what does that decision cost them?
Answering it takes the instruments of transportation engineering to measure, behavioural and decision science to explain, statistics and machine learning to generalise, and software engineering to deliver. I work across all four — and I build the instruments as well as the models, because a finding that never reaches a commuter or a planner has not finished its job.
Discrete choice, hybrid latent-class and integrated choice latent variable models; stated-preference design, willingness-to-pay estimation and two-wave panel field experiments.
Measuring what commuters actually inhale across transport micro-environments, inhaled-dose comparison, and health-impact assessment of active travel.
Deep and geostatistical models for sparse sensor networks, interpretable learning for choice prediction, computer vision for the built environment, and convex reformulations of structured choice estimation — AI held to the standard a behavioural model is held to.
Routing algorithms and decision-support tools, accessibility and transit-oriented development prioritisation, and evidence built for transport and clean-air policy.
This is the shape of the models I estimate, running in your browser. Change the scenario and watch the modal split re-solve: utilities, a softmax over the choice set, and the log‑sum that measures what the whole set is worth to a traveller.
Unhealthy
How heavily this traveller weighs what they breathe
Sets βtime against βcost
Predicted modal split
| Mode | Time | Cost | Inhaled PM2.5 | V | Share |
|---|
Illustrative. The structure — linear-in-parameters utility, a logit kernel, exposure entering as inhaled dose rather than ambient concentration — is the one I estimate and publish. The coefficients here are chosen to make the mechanics legible, and are not estimates from any of the papers below.
of a commuter’s daily pollution dose is inhaled during travel — which takes only 8% of the day.
exposure in Central Delhi when commuters take the least-exposure route, at a cost of +40% travel time.
energy consumption in South Delhi on the least-energy route — cleaner air and lower fuel cost together.
continuous air-quality stations operating in India against roughly 4,000 needed — why the models must interpolate.
citations on Google Scholar — h-index 6, i10-index 6.
study contexts across India — Delhi, Kolkata, Mumbai, Kharagpur and tier-2/3 cities.
The core of my dissertation is a two-wave route-choice panel: the same 723 Kolkata commuters surveyed once in winter and again in summer, yielding 5,224 choice observations. Following the same people across seasons separates a real change in preference from a change in who happens to answer.
Air quality weighs roughly four times more heavily on route choice in winter than in summer, and willingness to pay for cleaner air is 1.8× higher per AQI band. Commuters valued a pre-trip AQI alert more than an equivalent reduction in physical exposure — which makes real-time air-quality information the cheapest policy lever available. AQI literacy, not income, is the dominant moderator.
Panel binary logit, machine learning and deep learning models, triangulated.
Stated-preference panels and on-road sensing across transport micro-environments.
Hybrid latent-class and ICLV discrete choice models of who changes, and who cannot.
Deep and geostatistical models filling the gaps between sparse monitoring stations.
DRUM: a live routing engine returning the least-exposure and least-energy route.
Evidence for transport policy, TOD prioritisation and active-mobility investment.
One line of work — how commuters trade travel time against the air they breathe — became a paper, a routing engine people can use, a full spread in The Hindu, and now a patent.
The DRUM app, developed by IIT-Kharagpur researchers, has shown promise in simulations and will soon start real-world tests; its makers are also exploring integrating data from low-cost sensors on street poles and those carried by commuters.
“A major advantage of crowdsourced data is that it would allow us to expand the model beyond cars and two-wheelers, which are currently the only modes included.” Kapil Kumar Meena — The Hindu, June 2025
Research earns its keep when it reaches someone. This one took six years to get from a sensor on a Kharagpur roadside to an application published by the Indian Patent Office.
Patent application · Government of India
Filter by topic, type, year or venue. My name is set in bold.
Meena, K. K., Goswami, A. K.
Transportation Research Record2026
Meena, K. K., Goswami, A. K.
Transportation Research Record2026
Meena, K. K., Goswami, A. K.
Transportation Research Part A: Policy and Practice2025
Meena, K. K., Goswami, A. K.
Travel Behaviour & Society2025
Meena, K. K., Singh, A. K., Goswami, A. K.
Transportation Research Record2025
Sumbhate, A., Meena, K. K., Goswami, A. K.
Journal of the Eastern Asia Society for Transportation Studies, 162025
Manoj, B. S., Meena, K. K., Panchal, H., Sharma, G., Goswami, A. K.
International Journal of Sustainable Transportation2025
Manoj, B. S., Meena, K. K., Goswami, A. K.
Sustainable Transport and Livability, 22025
Meena, K. K., Goswami, A. K.
Transport Policy, 154, 48–602024
Meena, K. K., Bairwa, D., Agarwal, A.
Decision Analytics Journal, 11, 1004592024
Meena, K. K., Singh, V., Agarwal, A.
Journal of Transport & Health, 31, 1016432023
Singh, V., Meena, K. K., Agarwal, A.
Urban Climate, 38, 1009012021
Meena, K. K., Goswami, A. K.
Transportation Research Board (TRB) 105th Annual Meeting, Washington, DC2026
Gupta, C., Amitabh, A., Meena, K. K., Goswami, A. K.
18th International Conference on Communication Systems & Networks (COMSNETS), IEEE2026
Joshi, V., Meena, K. K., Goswami, A. K.
IEEE International Conference on AI Engineering and Innovation (AIEI)2026
Singh, A., Meena, K. K., Sharma, G., Goswami, A. K., Mishra, S.
Transportation Research Board (TRB) 105th Annual Meeting, Washington, DC · also WCTR, France2026
Meena, K. K., Singh, A. K., Goswami, A. K.
7th Conference of the Transportation Research Group of India (CTRG), SVNIT Surat2025
Sumbhate, A., Meena, K. K., Goswami, A. K.
Eastern Asia Society for Transportation Studies (EASTS), Jakarta2025
Manoj, B. S., Meena, K. K., Goswami, A. K.
1st World Symposium on Sustainable Transport and Livability (WSSTL), IISc Bengaluru2025
Kodukulla, R., Meena, K. K., Sharma, G., Goswami, A. K.
Transportation Infrastructure Projects: Conception to Execution (TIPCE), IIT Roorkee2025
Dasgupta, S., Meena, K. K., Majumdar, D., Goswami, A. K.
Energies, AEEE India Conference2025
Meena, K. K., Taneja, R., Agarwal, A.
16th International Conference on Communication Systems & Networks (COMSNETS), IEEE2024
Sumbhate, A., Meena, K. K., Goswami, A. K.
52nd Urban Affairs Association (UAA) Annual Meeting, Nashville2024
Mohanty, P., Meena, K. K., Goswami, A. K.
52nd Urban Affairs Association (UAA) Annual Meeting, Nashville2024
Manoj, B. S., Meena, K. K., Panchal, H., Sharma, G., Goswami, A. K.
17th International Association for Travel Behaviour Research (IATBR), Vienna2024
Meena, K. K., Goswami, A. K.
16th World Conference on Transport Research (WCTR), Montréal2023
Meena, K. K., Kumar, R., Goswami, A. K.
14th Conference on Transport Planning and Implementation Methodologies for Developing Countries (TPMDC), IIT Bombay2022
Meena, K. K., Goswami, A. K.
Indian Patent Application no. 2026310183792026
Nothing matches those filters.
Every co-author on this page, laid out by a spring–charge simulation. Node size is papers together; colour is the area we work in most. Pick anyone to filter the list.
Each of these began as a chapter or a dataset and ended as something that runs in a browser. Source is public.
Dynamic Routing for Urban green Mobility. Five routes — shortest, fastest, least exposure, least energy, and a balanced suggestion — from live CPCB and traffic data.
The DRUM codebase: routing backend, segment-wise pollution interpolation, and the commuter-facing front end.
Machine-learning forecasts of next-day fine particulate concentrations, published as an interactive global viewer.
Visualising and predicting cycling and motorcycle mode share across Indian cities.
The deployed stated-choice instrument behind the two-wave commuter panel — the data source for my dissertation.
A mobile-first prototype exploring how route alternatives should be presented so a commuter will actually pick the cleaner one.
Kota to Roorkee to Kharagpur to Los Angeles, with a detour through policy consulting.
Recognition for the work, the funding that moved it across borders, and the reviewing that keeps the field honest.
Exposure-aware routing, behavioural modelling, active mobility, or a paper you think I should read. Students curious about transportation research are welcome too.