Simulations Reveal the Hidden Traffic Threat to Brazil’s Newest Innovation District

On the western edge of Campinas, Brazil, a stretch of former farmland known as Fazenda Argentina is being transformed into the International Hub for Sustainable Development, or HIDS, an innovation district owned by the University of Campinas and designed to fuse laboratories, research institutes, and technology companies into a single knowledge-intensive urban environment. Because the site is still largely empty, researchers at UNICAMP saw a rare opportunity: they could stress-test the district’s planned road network before a single building opened. Using trip generation estimates drawn from the transportation literature and detailed microscopic traffic simulations, a team led by Vitor Eduardo Molina Junior set out to answer a question that haunts urban planners across the Global South—how do you plan mobility for a city district that does not yet exist, using data that has not yet been collected?

The study, published in Clean Technologies and Environmental Policy, arrives at a finding that upends conventional expectations. The biggest traffic threat to the new district is not the traffic the district itself will generate. It is the through traffic—vehicles from elsewhere that could hijack the campus’s shiny new roads as a shortcut between the Campinas–Mogi Mirim highway and the surrounding urban network. In the most demanding simulated scenarios, segments of Ricardo Benetton Avenue, the district’s main access corridor, exceeded their operational capacity, with volume-to-capacity ratios climbing above 1.0 and Level of Service degrading to the worst category, F, under the criteria of the Highway Capacity Manual. In plain terms, the road would choke not because of the scientists and entrepreneurs moving into the district, but because everyone else discovered it was the fastest way across town.

The methodology behind this conclusion is a careful exercise in scenario-based planning under uncertainty. Conventional transport planning relies on the four-step travel demand model—trip generation, trip distribution, mode choice, and trip assignment—which requires years of local socioeconomic and behavioral data to calibrate. For a greenfield site like HIDS, that data simply does not exist. The researchers therefore adapted the framework: they characterized the planned land uses for each implementation phase approved by UNICAMP’s University Council in September 2024, estimated trip volumes using published trip generation rates for research and development facilities, and treated the district as a multi-use trip generator, an activity cluster whose components share a common road network.

The land-use assumptions were themselves phased, mirroring the district’s occupation plan. Phase 1, running from 2025 to 2030, covers just 2.5 percent of the Fazenda Argentina site and includes 25,200 square meters of built area devoted to laboratories, research offices, and administrative facilities. Phase 2, from 2030 to 2040, adds another 30,000 square meters, while Phase 3, from 2040 to 2050, brings a further 84,200 square meters, including a metropolitan hospital. Because the planned uses are almost entirely research-oriented—unlike the mixed residential-commercial fabric of most urban developments—the researchers relied on rates calibrated for a comparable technology park in southern Brazil, supplemented by standard rates from the Institute of Transportation Engineers. They openly acknowledge a caveat: suburban-style ITE rates, derived from North American contexts, may overestimate vehicle demand in districts designed around public and active transport, an emerging concern in trip generation research across developing countries.

With demand estimated, the team turned to spatial modeling before simulation ever began. Geographic and semantic data from UNICAMP and the Campinas municipal geographic databases were used to build a parametric model in Grasshopper for Rhino 3D, refined into a comprehensive City Information Model in Esri CityEngine and later extended in Autodesk Revit to capture the road network beyond the district’s boundaries. The proposed street cross-sections varied deliberately: Ricardo Benetton Avenue with four 2.8-meter lanes, Health Avenue with two generous 5.0-meter lanes in each direction, and shared and collector streets with single 3.6-meter lanes. This digital twin of a not-yet-built district then became the testbed for the microsimulation engine Aimsun Next 20.0.5.

The sensitivity analysis was deliberately brute-force. Because no current traffic counts existed for the corridor, the researchers anchored their baseline at 500 vehicles per hour—an order-of-magnitude estimate supported by a traffic impact report for a nearby development and validated with preliminary counts from thermal cameras equipped with onboard artificial intelligence for vehicle detection. From that baseline, they ramped demand upward in increments of 500 vehicles per hour, running ten morning-peak scenarios from 500 to 5,000 vehicles per hour, each layered with the roughly 310 trips per hour expected from Phase 1 activities. The escalation revealed a clear threshold: conditions began deteriorating at 3,500 vehicles per hour, and at 4,000 vehicles per hour and above the corridor crossed into progressive degradation, with oversaturated segments and Level of Service collapse. Notably, the district’s own trips barely dented the operating conditions—the problem was the pass-through volume.

Having identified the breaking point, the researchers then asked whether public transportation could pull the corridor back from the edge. They modeled the bus routes currently serving the main UNICAMP campus, extended to serve HIDS, with stops spaced 200 meters apart in line with walkability principles. Two occupancy scenarios were tested—40 seated passengers and 70 passengers per bus—against the critical 4,000-vehicle-per-hour baseline, assuming an average private car occupancy of 1.2 people and that 80 percent of transit riders would alight within the study area. The arithmetic was striking: a fleet of 16 buses operating in the peak hour removed 534 vehicles from the road in the minimum-occupancy case and 934 in the maximum-occupancy case. In the high-occupancy scenario, volume-to-capacity ratios dropped below 1.0 on previously saturated segments, restoring workable—if still imperfect—conditions classified between Levels of Service C and E.

The implication is both practical and provocative. Rather than widening roads or building new lanes, Campinas could keep its new innovation district flowing simply by making buses frequent and full enough. Delay rates, it turned out, were largely locked in by the corridor’s geometry and conflict points, so the levers that mattered were demand-side: vehicle access control, restrictions on through movements, and policies that favor transit, cycling, and shared mobility. The simulations also carry a warning for the surrounding landscape. Adjacent agricultural land is already slated to become a gated, car-oriented residential development of more than 700 plots potentially housing over 10,000 people, and at least two neighboring farms show similar conversion potential. Campinas itself is expected to grow by only about 25,000 residents over the coming decade, so the district’s success will hinge on whether the city uses its zoning power to actively promote compact, walkable, transit-oriented growth rather than merely permitting it.

Beyond its local significance, the study fills a conspicuous gap in the literature on knowledge-based urban development. Systematic reviews have shown that research on innovation territories—everything from Triple Helix science parks to fourth-generation districts integrating civil society and environmental preservation—remains fragmented and geographically skewed, with peripheral and semiperipheral contexts underrepresented and urban mobility rarely treated as a measurable network problem. Silicon Valley, Barcelona’s 22@ District, and Seville’s Cartuja park have been dissected at length, but the operational traffic dynamics of emerging districts in the Global South, where infrastructure decisions must be made before mobility data exists, have gone largely unexamined. This study’s exploratory framework—linking phased land-use plans, trip generation rates, and HCM-based microsimulation—offers a replicable template for exactly those data-scarce settings.

The authors are careful to frame their results as a decision-support tool rather than a prophecy, and they stress that continuous monitoring and future traffic counts will be essential to calibrate the models as the district fills in. But the core message travels far beyond Campinas. Innovation districts are often marketed as sustainable urbanism, yet when they land on urban fringes dependent on automobile access, they can quietly become congestion engines unless land-use and mobility policies are integrated from day one. The HIDS experience suggests that the cheapest, most sustainable traffic infrastructure a new knowledge district can build is not asphalt at all—it is a bus system people actually choose to ride, delivered before the first parking space is ever painted.

Subject of Research: Traffic impact microsimulation of the HIDS-UNICAMP innovation district in Campinas, Brazil

Article Title: An exploratory traffic impact analysis in a knowledge-based urban development area: the HIDS/UNICAMP case study

Article References: Molina Junior, V. E., Emiliano, W. M., dos Santos Bardini, V. S., de Oliveira e Sousa, M. N. P., Cruz, S. R. S., Rodrigues, B. N., Favoreti, A. L. F., Telhada, J., & Canteras, F. B. (2026). An exploratory traffic impact analysis in a knowledge-based urban development area: the HIDS/UNICAMP case study. Clean Technologies and Environmental Policy, 28(10), Article 246. https://doi.org/10.1007/s10098-026-03592-8

Image Credits: AI Generated

DOI: 10.1007/s10098-026-03592-8

Keywords: innovation district, traffic microsimulation, knowledge-based urban development, trip generation, transit-oriented development, HIDS, UNICAMP, sustainable mobility, urban planning, Level of Service, public transportation, Global South