I often find myself standing on a busy curb in my city, watching the choreography of delivery vans, ride-hail drivers, cyclists, and pedestrians negotiate the same narrow strip of public space. As someone who follows smart-city tech closely, I keep asking: if we handed curb management to an AI, how would it decide who gets to stop where and when? Would it favour deliveries over pickups? Protect pedestrians? Or simply optimize for revenue?
What is AI-driven curb pricing?
AI-driven curb pricing is a system that uses algorithms, sensors and real-time data to dynamically set prices for curb access. Instead of static "loading zone" signage, prices and permissions could change by the minute to reflect demand, safety conditions, traffic flow and broader policy goals like reducing emissions or promoting active travel.
At Mobility News, I track platforms experimenting with dynamic kerb management — think of companies like Coord (now part of Mastercard), Remix (for planning) and startups that integrate camera feeds, payments and routing APIs. But the key question is not whether it can be done — it's how the AI makes trade-offs between competing users.
Which objectives would the AI optimize?
Before deciding between deliveries, pickups and pedestrians, the AI needs objectives. Those will be defined by policymakers and stakeholders, and they shape everything.
The AI doesn't "want" anything by itself — it will try to satisfy the weighted combination of these objectives. My curiosity is always about those weights: who sets them and how transparent they are.
What data would the system use?
An effective system needs rich, real-time data:
Combining these inputs, the AI can infer probable conflicts — for example, a sudden surge in deliveries in a shopping street during lunch — and adjust prices or permissions accordingly.
How would the AI assign priorities in real time?
Imagine a curb with three competing requests: a food-delivery van needs 10 minutes to load, a ride-hail car has a passenger arriving in two minutes, and a group of pedestrians is crossing to access a market. Here's how the AI might evaluate options:
So the decision is rarely binary. A well-designed system might split the window: allow the delivery to use a smaller slot, require rapid loading protocols, or temporarily reserve the curb for pedestrians during peak market times.
Would AI favour money over people?
This is the political heart of the debate. Algorithms can be tuned to prioritise revenue, but they can also embed safety and equity constraints. In my view, the danger is twofold:
I've seen pilots that intentionally discount curb access for cargo bikes and zero-emission vans to encourage greener last-mile solutions. That's an example of using pricing as a policy lever rather than a pure revenue tool.
How can pedestrians be protected?
Pedestrians are the most vulnerable and yet the least likely to generate direct revenue. Here are mechanisms to protect them:
From my reporting, cities like Barcelona and London are experimenting with 'pedestrianisation windows' where commercial access is limited. AI can make these windows more responsive, but public oversight remains essential.
What about fairness and accountability?
Fairness requires both technical safeguards and governance. I believe in three practical steps:
Without these, a curb-pricing system risks becoming a black box that privileges those who can pay or those plugged into well-resourced APIs.
What are realistic rollout scenarios?
Here are three paths I see in cities:
- Operator-driven pilots: Logistics firms and platforms trial dynamic curb payments in limited zones (e.g., ports or business districts).
- Regulated deployment: Municipalities set policy goals (safety-forward, equity-forward) and require vendors to conform to them.
- Hybrid marketplaces: A public API allows private actors to bid for curb slots, but with municipal vetoes for safety and equity concerns.
What happens to small businesses and gig workers?
Small retailers and couriers are especially vulnerable. To avoid disproportionate harm, cities can:
I've talked with local couriers who fear being priced out of prime drop-off zones. Any practical system must include transition measures to protect livelihoods.
How will we know if it's working?
| Metric | Why it matters |
| Average curb occupancy time | Shows if turnover is improving |
| Pedestrian injury incidents | Tracks safety impacts |
| Delivery punctuality for essential goods | Assesses service quality |
| Distribution of fees paid by user type | Monitors equity |
| Mode share (cargo bikes, EVs vs ICE vans) | Measures environmental effects |
Transparent dashboards — updated in near real-time — help build trust. I want to see cities publish these metrics, not just the aggregate revenue figures.
At the end of the day, AI-driven curb pricing is a powerful tool, but it's not a neutral arbiter. The technical choices and policy priorities embedded in the system will determine whether curbs become fairer, safer and greener, or whether they merely become another monetised layer that shifts burdens onto the less powerful. As someone who cares about equitable mobility, I'm optimistic that with the right governance, we can design systems that benefit people first — and use dynamic pricing as a means to achieve shared urban goals rather than an end in itself.
For more analysis and updates on curb management pilots, delivery innovations and smart-city governance, visit Mobility News.
