AI traffic planning has spent two decades being reactive. A queue forms, a camera sees it, a control room retimes a signal, and the delay that already happened is shaved by a few seconds. A team at New York University’s Tandon School of Engineering has published a framework that tries to move the work upstream, forecasting where congestion will sit years out and letting a planner interrogate the result in plain English.
The paper, “Geospatial AI Applications for Reducing Traffic Congestion and Guiding Planning Decisions,” appears in Transactions in GIS, volume 30, issue 4, under DOI 10.1111/tgis.70327. Anton Rozhkov of NYU’s Center for Urban Science and Progress wrote it with Pranav Nitin Motarwar and Rudra Patil. Our AI models and tools hub tracks the model families this kind of applied work is built on, and the predictive analytics page covers the forecasting side in a business context.
This article sets out exactly what the framework does, what the numbers in it mean, where the arithmetic leads, and what the authors say it cannot yet do. The result is narrower than the headlines suggest and more useful because of it.
Table of contents
- What the AI Traffic Planning Framework Actually Is
- The Numbers Behind the AI Traffic Planning Claim
- Why Two Models Instead of One in AI Traffic Planning
- The Spatial Half of AI Traffic Planning Matters More Than the Forecast
- Where AI Traffic Planning Sits Against What Cities Run Today
- The Query Portal Is the Part Most Cities Would Notice
- What the AI Traffic Planning Study Does Not Establish
- How to Read an AI Traffic Planning Result If You Buy Technology
- What Happens Next for AI Traffic Planning at NYU
- Frequently Asked Questions About AI Traffic Planning
- References
What the AI Traffic Planning Framework Actually Is
The AI traffic planning system is not one model. It is four components wired into a single workflow, and the AI traffic planning value comes from the join rather than from any one part.
A forecasting layer with two competing models
The framework runs ARIMA, a conventional statistical time-series method, alongside an LSTM neural network, and compares them on the same held-out data. Keeping both is deliberate: the comparison is the evidence that the neural approach earns its complexity.
A spatial layer built on hexagons
AI traffic planning forecasts are laid onto Uber’s H3 hexagonal grid, which tiles the city into cells that can be zoomed in or out without changing shape. That lets the same data answer “which borough” and “which corridor” without a separate model for each question.
A clustering step that finds hotspots
On top of the hexagons, clustering groups cells that behave alike, which is how the framework locates congestion hotspots rather than reporting one flat number for the whole city.
A language interface grounded in the city’s own data
The last component is a customised portal built on Meta’s LLaMA model, connected to a project-specific traffic knowledge base and run zero-shot. Planners type a question; the portal answers from the project’s data rather than from whatever the base model absorbed in training. This is applied natural language processing doing an unglamorous job well.
Why it runs locally
Rozhkov’s stated motivation for AI traffic planning tooling was control over where the data sits. “We started with an idea: what if we developed our own AI platform, one that could be hosted locally and would be secure, intuitive and comfortable for planners to use in their day-to-day work,” he said — a platform agencies can run “behind their own firewall.”
| Component | What it contributes |
|---|---|
| ARIMA | Statistical baseline; captures repeating seasonal structure |
| LSTM | Neural forecaster; captures non-repeating pattern |
| H3 hexagonal grid | Multi-scale spatial indexing of forecasts |
| Clustering | Groups cells into congestion hotspots |
| LLaMA query portal | Plain-language access, grounded, zero-shot |
| Deployment | Local hosting, behind an agency firewall |
The Numbers Behind the AI Traffic Planning Claim
Two figures carry the paper, and both are worth reading slowly before deciding what AI traffic planning has actually proved here.
Fifteen years in, four years out
The AI traffic planning framework was fitted on roughly 15 years of New York City traffic observations spanning 2009 to 2024, and evaluated on a 2021 to 2024 test set. Four years of testing against fifteen years of record is about 27% of the data held back, which is a generous split by forecasting standards.
The error gap, stated precisely
On that test set the LSTM returned a root mean square error of 342.56 vehicles per day. ARIMA returned 417.62. The difference is 75.06 vehicles per day, which is 17.97% of the ARIMA figure — the “roughly 18% improvement” the coverage quotes, and it checks out.
What that error looks like in context
Against the framework’s own 2025 baseline of 12,540 vehicles per day, an error of 342.56 is 2.73% of daily volume. ARIMA’s 417.62 is 3.33%. Both are small; the neural model is better by about six tenths of a percentage point of daily volume, not by an order of magnitude.
The forecast itself
The AI traffic planning framework projects average daily traffic volume rising from 12,540 vehicles in 2025 to 19,680 in 2029. That is 7,140 additional vehicles per day across four years, an average of 1,785 per year, and a 56.9% total increase.
The growth rate hiding inside it
Compounded rather than averaged, 12,540 to 19,680 over four years is about 11.9% per year. Treating that as a certainty would be a mistake — the researchers stress the forecasts carry substantial prediction intervals — but it is the number a capital plan would have to absorb.
Why Two Models Instead of One in AI Traffic Planning
The ARIMA-versus-LSTM comparison is the part of the study most likely to transfer to other cities, because it is a statement about traffic data rather than about AI traffic planning fashion rather than about New York.
Seasonality is real but partial
Motarwar’s summary of the baseline is blunt: “ARIMA gives you seasonality, which is true but not the whole story.” Weekly and annual rhythms are genuinely there, and a statistical model finds them cheaply.
The residual is where the gain lives
“The LSTM picks up the parts of the pattern that don’t repeat cleanly, and that’s where most of the improvement came from,” Motarwar said. That is an unusually precise attribution — the 18% is not diffuse model magic, it is the non-repeating remainder.
Which argues against replacing ARIMA
If the improvement lives in the residual, the statistical model is still doing most of the work on the bulk of the signal. Running both is not indecision; it is how you know which part of the forecast you should trust least.
The comparison is also the honesty check
A single-model paper reporting 342.56 vehicles per day of error would give a reader nothing to judge it against. The paired result is what makes the AI traffic planning claim falsifiable by the next team that tries it.
Where the approach is likely to travel
Any city with a long, dense count record and a mix of repeating and irregular demand has the same structure. Cities with short records, or with a single dominant commuting pattern, would likely see a smaller spread between the two models.
The Spatial Half of AI Traffic Planning Matters More Than the Forecast
A forecast of a citywide average is close to useless operationally in AI traffic planning, and the paper’s authors say so directly.
Averages hide the problem
“A citywide average doesn’t help the city planners,” Patil said. A single number rising by 56.9% tells an agency that something is coming, and nothing about where to put the money.
Corridors, not boroughs
“What they need to know is which corridors are badly impacted, and those turn out to be consistent year after year,” Patil said. Persistence is the operationally valuable finding: a hotspot that recurs is a hotspot you can plan against.
Hexagons make scale a dial, not a rebuild
H3 cells nest, so the same indexed forecast answers a borough-level question and a corridor-level question without re-modelling. That is a plumbing detail with real consequences for how often a planning team can actually ask something.
What the hotspot analysis surfaced
Manhattan came out as the highest-congestion borough, with Brooklyn and Queens also appearing in the analysis, and specific concentrations around Lower Manhattan and the approaches near LaGuardia Airport. None of that will surprise a New Yorker, which is the point — it is a sanity check on the method.
Consistency is the testable claim
If the same corridors recur year after year, then next year’s data is a live test. That makes this spatial layer of AI traffic planning easier to validate than the five-year volume forecast, which nobody can check until 2029.
| Figure | Value | Source |
|---|---|---|
| LSTM test error | 342.56 vehicles/day | Published |
| ARIMA test error | 417.62 vehicles/day | Published |
| Absolute gap | 75.06 vehicles/day | Subtraction |
| Relative gap | 17.97% | 75.06 / 417.62 |
| 2025 baseline volume | 12,540 vehicles/day | Published |
| 2029 forecast volume | 19,680 vehicles/day | Published |
| Four-year increase | 7,140 vehicles/day, +56.9% | Subtraction, division |
| Implied annual growth | ≈11.9% compounded | Fourth root of 1.569 |
Where AI Traffic Planning Sits Against What Cities Run Today
Most large agencies already own forecasting tools. Understanding what is genuinely new here means being precise about what those tools do badly.
Adaptive signal control is reactive by design
Systems that retime signals from live detector feeds are the current state of practice, and they work. They also start from a queue that already exists, which is the ceiling this AI traffic planning work is trying to lift.
Four-step travel demand models are slow to question
The classic regional demand model is powerful and expensive to run. A scenario takes weeks, which quietly limits how many scenarios get considered before a decision is made.
Navigation apps optimise for the driver, not the network
Commercial routing data tells a city where delay happened yesterday from the point of view of individual drivers. It does not give an agency a structural account of why that corridor fails.
Congestion pricing needs a forecast it can defend
Any charging scheme has to withstand a public inquiry. A published error figure on a held-out test set is the sort of thing that survives that process, which is one practical reason the AI traffic planning literature is moving toward stated baselines.
The gap this fills is speed of enquiry
None of the above is replaced. What changes is that a planner can ask a spatial question and get an answer in minutes, which is the difference between testing one option and testing eight.
The Query Portal Is the Part Most Cities Would Notice
The forecasting half is standard AI traffic planning practice done carefully. The interface is the part that changes who can use the output.
The failure mode it is built against
“A general chatbot gives you a reasonable-sounding generic paragraph about congestion,” Rozhkov said. That sentence describes the exact risk of bolting a commercial assistant onto a planning workflow: fluent output with no connection to the agency’s own counts.
Grounding is the whole mechanism
The portal is tied to a project-specific traffic knowledge base. “Planners need an answer that comes from their own data,” Motarwar said. Grounding turns the model from an author into a retriever with a readable voice.
Zero-shot, deliberately
Running the portal zero-shot means no task-specific fine-tuning was needed, which lowers the maintenance burden for an agency that does not have a machine learning team on staff.
Local hosting solves a procurement problem, not just a privacy one
Transport agencies hold movement data that is sensitive in aggregate and politically awkward in detail. A platform that runs inside the firewall sidesteps a data-sharing review that can take longer than the modelling.
It lowers the cost of asking a second question
The practical effect of good AI traffic planning tooling is not one brilliant answer. It is that the fifth follow-up question costs the same as the first, so scenarios get explored rather than commissioned.
What the AI Traffic Planning Study Does Not Establish
The authors are explicit about the limits of the AI traffic planning work, and the limits are the most important paragraph in any applied paper.
It has not been trialled at scale
The AI traffic planning framework has not been through large-scale trials with actual planning departments. Everything above is a research result, not a deployment record.
It cannot reduce congestion by itself
The researchers say plainly that the platform cannot solve congestion on its own. A forecast changes what a capital programme knows; it does not add road capacity or move a single vehicle.
The forecast intervals are wide
The 2029 figure carries substantial prediction intervals. Quoting 19,680 as a point estimate, as most coverage has, drops the uncertainty that the authors attached to it.
One city, one data regime
Every result here is New York’s. A framework fitted on fifteen years of dense counts in a dense city says nothing yet about a mid-size city with sparse sensors.
A grounded model still inherits its base model’s habits
Grounding constrains what the portal retrieves. It does not eliminate the possibility of a confidently worded synthesis that overstates what the underlying cells support.
How to Read an AI Traffic Planning Result If You Buy Technology
For anyone evaluating vendor claims in this space, the paper is a useful AI traffic planning yardstick precisely because it is modest.
Ask for the baseline, not the headline
An 18% improvement is only meaningful next to the thing it improved on. A vendor quoting an accuracy number with no stated baseline has told you nothing.
Ask what the error means in units you use
342.56 vehicles per day is interpretable. “97% accurate” is not, unless you know accurate at what and against which alternative.
Ask where the model runs
Local hosting is a genuine differentiator for public bodies, and it is testable at procurement rather than after deployment.
Ask what the system refuses to answer
A grounded portal should decline questions its knowledge base cannot support. A system that always has an answer is the failure mode Rozhkov described.
Ask when the forecast becomes checkable
The corridor-persistence claim is testable next year. The volume forecast is not testable until 2029. Buy against the first.
| Claim | Status |
|---|---|
| LSTM beats ARIMA on this data | Measured on a held-out test set |
| Hotspots recur year to year | Stated by the authors, checkable |
| Volume rises 56.9% by 2029 | Forecast with wide intervals |
| Planners can query in plain English | Built and described, not trialled at scale |
| The approach reduces congestion | Explicitly not claimed |
What Happens Next for AI Traffic Planning at NYU
The paper carries 69 references, which is a signal about how much prior work it is standing on rather than displacing.
The obvious next step is a live pilot
Everything the authors flag as missing points the same way: an agency running the portal against its own live feed, with planners who did not build it deciding whether the answers are useful.
Corridor persistence is the cheapest thing to validate
A single additional year of counts tests whether the hotspot list holds. That is a far smaller commitment than validating a five-year volume forecast and it would move the AI traffic planning claim from plausible to evidenced.
The forecasting half is portable, the knowledge base is not
ARIMA and LSTM transfer to any city with a count record. The grounded query portal has to be rebuilt around each city’s own documents, which is where the real integration cost sits.
Wider prediction intervals would help, not hurt
Publishing the intervals alongside the point estimates would make the work harder to over-quote, and the coverage so far shows exactly why that matters.
The framing is the contribution
Stripped of the numbers, the argument is that AI traffic planning should be judged on whether a planner can ask a follow-up question, not on whether a model shaved a few vehicles off an error metric. That is a better standard than the field usually applies.
Frequently Asked Questions About AI Traffic Planning
What is the new AI traffic planning framework?
It is a geospatial AI system from NYU Tandon that combines ARIMA and LSTM forecasting, H3 hexagonal spatial analysis, clustering for hotspot detection, and a locally hosted LLaMA query portal grounded in a city-specific traffic knowledge base.
Who built it and where was it published?
Anton Rozhkov, Pranav Nitin Motarwar and Rudra Patil of New York University’s Tandon School of Engineering, in Transactions in GIS, volume 30, issue 4, DOI 10.1111/tgis.70327.
How much better was the neural model?
Root mean square error of 342.56 vehicles per day for the LSTM against 417.62 for ARIMA on the 2021 to 2024 test set — a gap of 75.06, or 17.97%.
What does it forecast for New York?
Average daily traffic volume rising from 12,540 vehicles in 2025 to 19,680 in 2029, an increase of 7,140 per day or 56.9%, with substantial prediction intervals attached.
Why does the LLaMA portal run locally?
So agencies can keep sensitive movement data behind their own firewall, and so answers come from the project’s own traffic knowledge base rather than a general model’s training data.
Has any city deployed it?
No. The authors state the framework has not undergone large-scale trials with actual planning departments.
Does AI traffic planning reduce congestion on its own?
No, and the researchers say so. It changes what planners know and when they know it; capacity, pricing and scheduling decisions still have to follow.
References
More AI coverage: explore Progressive Robot's AI Models, Tools & Releases hub — hands-on reviews, setup guides and benchmarks in one place.
