The most expensive question in urban planning is 'what if.' What if we close this lane, add that bridge, or change the bus schedule? Mobiliti Labs, a new startup from UC Berkeley, is building a platform to answer those questions with a simulation, not a spreadsheet. Its core technology is a massive-scale spatial intelligence simulator, licensed from Lawrence Berkeley National Laboratory and UC Berkeley, designed to model traffic and infrastructure changes in real time using high-performance computing and AI [ITS Berkeley, Unknown].
The Academic Wedge into Govtech
Founder Dr. Jane Macfarlane brings a pedigree that is rare for a pre-seed company. She is the Director of the Smart Cities Research Center at UC Berkeley's Institute of Transportation Studies and holds a joint appointment at Lawrence Berkeley National Laboratory [ITS Berkeley, LBNL]. This isn't a founder with a fresh idea looking for a technical co-founder, it's a researcher commercializing her lab's work. The licensed software, called Mobiliti, uses Parallel Discrete Event Simulation (PDES) and parallel traffic assignment to generate vehicle-level metrics for entire metropolitan areas, a task that traditionally required supercomputing clusters or relied on highly aggregated, less precise models [ITS Berkeley, Unknown]. The company's bet is that this academic-grade engine, packaged as a SaaS platform, can become the standard tool for city and state transportation agencies.
For a sector known for long procurement cycles and risk aversion, the product's origin story is a potential shortcut. The technology has been developed and validated within a national lab context, which carries inherent credibility with public-sector buyers. The initial target is clear: city planners, state DOTs, and organizations that manage large-scale infrastructure [Mobiliti Labs, 2025]. The value proposition is reducing the financial and political risk of multi-million dollar projects by simulating outcomes before breaking ground.
The Technical Breakdown
The platform's architecture reveals its ambitions. It isn't just another traffic visualization tool. According to research publications, the underlying Mobiliti engine is built for concurrency and scale [Lawrence Berkeley National Laboratory, 2019].
- Parallel Discrete Event Simulation (PDES). This is a core technique for distributing a simulation across many processors, allowing it to model millions of individual vehicle trips simultaneously rather than as statistical aggregates. This enables granular, second-by-second analysis of congestion and flow.
- AI/ML for forecasting. The system incorporates machine learning to improve traffic forecasting, feeding real-world data back into the simulation models to increase their predictive accuracy over time.
- Vehicle-level metrics. The output isn't just average speed or volume. It can theoretically show the impact of a policy change on specific neighborhoods or demographic groups, a key consideration for equity-focused planning [Institute of Transportation Studies, UC Berkeley, Unknown].
This technical foundation is the company's primary asset. The challenge will be productizing it for users who are experts in civil engineering, not distributed systems.
The Scale Question
While the technical credentials are strong, Mobiliti Labs faces the classic spinout challenge of moving from research to revenue. The public record shows no disclosed funding, customers, or deployments. The company is in a clear pre-revenue, stealth-building phase, operating with a waitlist for early access [Mobiliti Labs, 2025]. The path to scale in govtech is notoriously difficult, characterized by long sales cycles, complex procurement rules, and entrenched incumbents offering legacy planning software.
The sober assessment of what could go wrong hinges on two operational hurdles. First, the computational cost of running city-scale, high-fidelity simulations could make pricing prohibitive for all but the largest municipal budgets unless the company achieves significant efficiency gains. Second, the 'last mile' of product design, translating raw simulation output into actionable, plain-language reports for non-technical decision-makers, is an unsolved problem for the category. A powerful engine is only valuable if the person approving the budget can understand its output.
The company's most plausible answer to these risks is its founder's deep domain network and the validated IP. Dr. Macfarlane's position within the transportation research community provides a natural channel for initial pilot projects and feedback. If Mobiliti Labs can convert that academic credibility into a handful of flagship city deployments, it will have the case studies needed to tackle the broader market. For now, it represents a technically serious attempt to bring a new class of computational tool to an old problem.
Sources
- [ITS Berkeley, Unknown] Macfarlane, Mobiliti Labs License Innovative Software | https://its.berkeley.edu/news/macfarlane-mobiliti-labs-license-innovative-software
- [Lawrence Berkeley National Laboratory, 2019] Mobiliti: A shift for Analyzing Traffic Congestion | https://cs.lbl.gov/news-media/news/2019/mobiliti-a-shift-for-analyzing-traffic-congestion/
- [Institute of Transportation Studies, UC Berkeley, Unknown] Mobiliti,A New Tool to Guide Safer, More Equitable Traffic Management Strategies | https://its.berkeley.edu/publications/mobiliti,a-new-tool-guide-safer-more-equitable-traffic-management-strategies
- [Mobiliti Labs, 2025] Homepage | https://mobilitilabs.ai/
- [Mobiliti Labs, 2025] About Us | https://www.mobilitilabs.ai/about-us