For nine Palm Beach County high school students, West Palm Beach became more than the setting for a summer research program. It became part of the research itself.
Over six weeks, students in the 2026 My Streetscape Summer Research Institute explored how artificial intelligence, sensing and other emerging technologies could address challenges affecting people in an urban environment. Working in teams alongside researchers from the Institute for Smarter Cities, Spaces, and Health (I-SMART) at Florida Atlantic University (FAU), the students investigated issues ranging from pedestrian and railroad safety to sidewalk conditions, government transparency and civic engagement.
The program, offered through the NSF Engineering Research Center for Smart Streetscapes (CS3), introduces high school students to engineering and research through problems grounded in the communities around them. Rather than beginning with technology and searching for an application, students were encouraged to consider the people who use city streets and public spaces and the challenges technology might help address.
Students spent time in downtown West Palm Beach observing the built environment, examining how people interact with streets and infrastructure and considering the role technology already plays in public spaces. During a walking tour of downtown, for example, students studied crosswalks, building design, public art, shade and the placement of sensing technologies. By attending a public meeting at City Hall, students experienced how residents can participate in democratic processes, thus shaping local streetscapes. They also discussed how pedestrian data can help researchers and urban planners understand how people move through a city while raising important questions about privacy and responsible data collection.
The experience reflects an ongoing partnership among FAU, CS3 and the West Palm Beach Community Redevelopment Agency (CRA), which helps connect student research with the city and community where many of those questions take shape.
From Community Questions to Research Projects
As the summer progressed, students translated what they were learning about cities and their residents into five research projects. While the projects addressed different challenges, a common theme emerged: how technology can give people greater agency in the places where they live, travel and interact with government.
Sabrina Kaiser explored that question through “CitySpeak: An AI-Based Mascot for Conversational Civic Engagement”, a prototype designed to make it easier for residents to interact with local government. The project addresses barriers that can prevent people from participating in traditional civic processes. In interviews with eleven West Palm Beach residents, eight reported that they had never attended a city council or neighborhood meeting.
CitySpeak combines conversational AI with computer vision and access to municipal information. The prototype is designed to recognize when pedestrians enter an interaction area, invite them to participate in a conversation and provide useful information about their community. It can also capture feedback and organize it into summaries that could help officials better understand residents’ concerns.
The goal extends beyond creating another digital information kiosk. CitySpeak is intended to make civic participation more convenient and engaging while increasing citizen agency and representation and creating another avenue for dialogue between municipalities and their residents. Although developed with West Palm Beach in mind, the students’ concept could eventually be adapted for other communities.
Access to government information was also at the center of “Digital Transparency Initiative: An Intelligent Government Record Recommendation System for Citizen Access”, developed by Levni Apraku-Mensah.
The project examined a deceptively complicated question: How can residents better understand and access information held by their local government while still accounting for privacy, legal and ethical concerns? Apraku-Mensah developed a prototype website that helps users identify records relevant to their needs. Features included conversational assistance, a record recommendation system and tools for masking personally identifiable information.
Testing showed that the prototype performed correctly in four of five scenarios, while also revealing opportunities for improvement. The work points toward future collaboration with local government agencies, expanded educational resources and additional tools that could help residents better navigate public-record processes.
Exploring Safer Streets Through AI
Three of the summer’s projects focused on another fundamental part of city life: moving safely through public spaces.
Ryan Calixte and Cassandra Glomann developed “SmartWalk: AI-Based Pedestrian Detection and Driver Alert System for Intersection Safety” to explore how artificial intelligence could improve awareness between pedestrians and motorists at intersections.
Their system uses computer vision to detect pedestrians and vehicles, identify crosswalks, analyze pedestrian movement and estimate whether someone intends to cross. That information could ultimately be relayed to drivers, increasing awareness of pedestrians who may be difficult to see or whose movements may be difficult to anticipate.
The students compared several AI models as they developed different components of the system, learning that the newest or most complex model was not automatically the best choice for every task. Their results also exposed some of the challenges involved in moving an AI system from an experiment toward a real-world application, including limited training data and the risk of inaccurate predictions when some groups are underrepresented in the data.
Importantly, SmartWalk was designed around more than technical performance. Its goals included increasing pedestrian confidence and independence while protecting personally identifiable information. Future improvements proposed by the team include face-blurring technology and expanding the system to account for cyclists, animals and other road users.
Just a few blocks can separate an intersection from a railroad crossing, but the safety questions can be very different. Adrian Garcia and Gerald Mano tackled one of those challenges with “ProxiRail: A Retrospective Risk-Detection System for Identifying Near-misses”.
Traditional transportation safety data often captures collisions after they occur. Garcia and Mano instead asked whether computer vision could help researchers identify dangerous encounters that almost resulted in an accident.
The students developed a system that detects and tracks people and trains at railroad crossings, determines when crossing gates are active and records when pedestrians leave the tracks relative to an approaching train. By calculating the time between those events, the system can begin to quantify near-misses that otherwise might go undocumented.
Their work demonstrated that computer vision could be used to identify track conditions with high precision and provided a foundation for a system that could eventually catalogue near-miss events, identify particularly dangerous crossings and help researchers understand characteristics associated with increased risk.
Brian Li, Vanessa Pierre, and Daventzky Edouard approached pedestrian mobility from another perspective with “StreetSense: ML-Powered Sidewalk Monitoring.”
Sidewalk cracks, displacement and other physical problems can create barriers for anyone traveling on foot, but they can be particularly significant for people with mobility challenges and other vulnerable road users. StreetSense explored how machine learning and computer vision could make those problems easier to identify, organize and report.
The team designed a system that analyzes images for different types of sidewalk damage, ranks problems by severity and connects that information to a map-based dashboard. Their testing revealed an important research lesson: no single AI model performed best across all types of sidewalk defects. Instead, the students found that combining two models could reduce the chance that certain types of damage would go undetected.
The project was designed to complement, rather than replace, existing city services by giving pedestrians a clearer way to identify and communicate infrastructure problems. Future versions could expand beyond sidewalks to roads and bike lanes, eventually becoming a more comprehensive application for pedestrians.
Learning How Research Reaches Others
Developing the projects was only part of the summer experience. Students also learned that research must be communicated effectively to make an impact.
As their projects neared completion, representatives from FAU’s Office of Undergraduate Research and Inquiry (OURI) worked with the students to create effective research posters and explain technical work to different audiences. Students learned how to organize complex ideas, determine which results were most important and use visuals to make their research easier to understand.
Those skills were put into practice during final program presentations in West Palm Beach and again when the students traveled to FAU’s Boca Raton campus to present their posters at the OURI Summer Student Showcase. The experience gave students an opportunity to move through a research process that extended from identifying a problem and testing potential solutions to defending their choices, discussing limitations and sharing results.
Their posters also demonstrate that research rarely ends with a finished prototype. Each team identified limitations and future directions for its work, whether that meant improving safeguards for civic AI, expanding pedestrian detection systems, collecting stronger datasets or developing new ways to communicate infrastructure and safety information. That iterative process — learning from what worked, understanding what did not and determining what should come next — is fundamental to research.
A City as a Classroom
Taken together, the five projects show the breadth of questions that can emerge when students are asked to look closely at the places around them.
A conversation with local government. A damaged sidewalk. A busy crosswalk. A railroad gate. Each represents an everyday interaction with a city, but each can also become a research question when students are given the tools and mentorship to investigate it.
For the nine students in the 2026 My Streetscape Summer Research Institute, that meant learning to use artificial intelligence and other technologies while also thinking critically about whom those technologies serve. Across the projects, students repeatedly confronted questions involving privacy, accessibility, representation, safety and trust alongside the technical challenges of developing systems that work reliably.
The program’s connection to West Palm Beach makes that experience possible. FAU gratefully recognizes the West Palm Beach Community Redevelopment Agency for its continued partnership and support of the My Streetscape Summer Research Institute. Through this ongoing relationship, students have an opportunity to engage with a real urban environment and explore research questions grounded in the needs and experiences of the surrounding community.
The summer concluded with five projects and nine student researchers, but the larger outcome was an introduction to what research can look like when engineering, technology and community needs come together — and an opportunity for students to begin seeing themselves as people capable of contributing to that work.
Acknowledgment
This work was supported through the NSF Engineering Research Center for Smart Streetscapes (CS3), funded through NSF Cooperative Agreement EEC-2133516.