When a city or operator considers launching a new shared mobility service, the first instinct is often to “do the math” with familiar tools:
- Spreadsheets to calculate projected ridership, revenue, and costs.
- BI dashboards that aggregate historic trip data into averages and trend lines.
- Consultants’ demand forecasts built with regression models or four-step transportation planning methods.
These tools are great for quick estimates such as trip counts, revenue projections, or cost breakdowns.
But here’s the problem: these methods assume a stable world. They struggle when real-world dynamics kick in – weather shocks, new bike lanes, policy changes, or competing services.
But mobility systems are living ecosystems where demand, supply, infrastructure, and policy constantly interact.
The Limits of Traditional Modeling
- Too static: Most tools rely on averages and fixed assumptions. They cannot tell you how demand shifts on a rainy day, or how a new low-emission zone changes usage.
- One scenario at a time: A spreadsheet can compare “400 vs. 500 scooters,” but it’s impractical to test dozens of combinations across weekdays vs. weekends, good vs. bad weather, or new parking regulations. As a result, many edge cases – which often matter most – go untested.
- Partial vision: Demand, fleet operations, and infrastructure are often modeled separately. In reality, demand shifts affect vehicle availability, which in turn impacts charging and rebalancing needs – a web of interactions traditional tools can’t capture.
- Weak geospatial modelling: Spreadsheets, consultants’ forecast or even four-step models rarely go below the neighborhood level. But shared mobility operates street by street: whether a scooter is 200 m or 600 m from a user can decide if a trip happens at all.
- One-off outputs: Consultant reports are static snapshots; spreadsheets only update when someone manually changes the formulas. They don’t improve automatically as new trip or operations data comes in.
That might be fine for a quick back-of-envelope calculation. But when you’re making big decisions on fleet purchases, charging stations, or pricing strategies, the margin for error is too thin.
The Simulation Advantage
This is where SWITCH’s simulation engine, which powers Urbiverse, becomes essential.
Instead of just crunching numbers, it builds a digital twin of the city:
- Context Model: digital representation of any city in the world with data like land use, weather, events, and regulations.
- Demand Model: trained on trips, traffic flows and public transport across different city areas,
- Supply Model: a virtual replica of your fleet, infrastructure, and operational rules.
- Simulator: micro-simulations that test how demand and supply interact under different strategies.
The result? Not just point estimates, but a range of possible outcomes – and the KPIs you actually need to decide: coverage, unmet demand, revenue vs. cost, CO₂ reduction, service equity.
Why It Matters
Imagine City A is considering launching a new e-scooter service.
Traditional methodology (without simulation):
- The city hires consultants who build a spreadsheet-based demand forecast.
- They use population density by district, average trip length, and benchmark usage from other cities.
- The model outputs:
- “500 scooters will generate around 1,000 trips/day, bringing €1.2M annual revenue.”
- Based on this, the city approves a 500-scooter pilot.
But here’s what this approach misses:
- It treats the city as broad zones – ignoring that in some neighborhoods scooters are a 2-minute walk away, while in others they’re 15 minutes away (a deal-breaker for riders).
- It assumes average weather – but in City A, 20% of days are rainy, when demand drops sharply.
- It ignores operational constraints like battery swaps, parking compliance, or rebalancing costs.
- It provides a single outcome – one number for “expected trips” – with no insight into what happens if usage patterns shift.
Simulation methodology (with Urbiverse):
- Instead of averages, a digital twin of City A is built: street network, land use, weather, events, and transport connections.
- Demand is simulated at the trip level: who rides, where, and under what conditions.
- The model tests multiple strategies automatically:
- 400 vs. 500 vs. 800 scooters
- weekdays vs. weekends
- rainy vs. sunny days
- with/without new parking hubs in the city center
- Outputs are decision-ready KPIs: coverage, unmet demand, rebalancing costs, CO₂ savings, service equity.
The result?
- The spreadsheet says: “500 scooters = 1,000 trips/day.”
- The simulation shows:
- With 500 scooters, central districts are oversupplied while outer neighborhoods face unmet demand.
- Adding parking hubs near metro stations increases ridership by 15% and reduces rebalancing costs by 20%.
- On rainy days, trips drop by 30% – but demand for public transport integration rises.
That’s the difference between a static estimate and a living decision lab: one tells you “what might happen on paper,” the other shows you how the system behaves in the real world under different futures.
The Bottom Line
Spreadsheets and dashboards are calculators.
A simulation engine is a decision lab.
For cities and operators, that means the confidence to launch new schemes knowing not only what’s likely to happen – but also how different futures could play out.
Book a call and start your simulations
FAQs
Q: How does Urbiverse’s simulation engine go beyond prediction? A: It doesn’t just forecast demand – it recreates how the entire mobility ecosystem behaves. Demand, supply, and infrastructure are simulated together, showing how operational choices ripple across the network.
Q: Do cities need extensive proprietary data to start using it? A: No. Urbiverse can build a realistic digital twin using open and public data such as demographics, land use, GTFS feeds, and weather records. When operators share trip or fleet data, accuracy simply increases.
Q: What insights can decision-makers expect? A: Actionable KPIs like coverage, unmet demand, revenue, rebalancing costs, and emissions – presented not as single values, but as ranges across different “what-if” scenarios, helping teams choose the best strategy.
Q: How often can the simulations be updated? A: Continuously. The model learns from new operational or trip data, so simulations improve as conditions evolve – unlike static consultant reports that freeze after delivery.
Q:Who benefits most from using Urbiverse?
A: Urban mobility operators, infrastructure investors, and urban planners who need to test fleet sizes, pricing models, or station placement before launching a service — and later use the same simulations to review, adjust, and optimize their strategy once the service is live.
