Team Smart Campus

<p>The video presents Team Data Exploit’s work on the "Formatting Machine" project, focusing on AI-driven tools for generating educational and marketing materials. Key modules demonstrated include a text-to-image application (using services like DALL-E and LightLLM) that transforms prompts into images, and an image-to-video tool (via Runway) that animates images based on user input. Both tools offer features for customizing style, size, quality, and AI-enhanced prompts, though current API limitations—cost, video length, and input restrictions—are noted.</p>
<p>The presenter discusses the rapid development of generative AI, challenges in scalability, and opportunities for integrating these modules into educational content (e.g., with avatars and slides). The technology is seen as highly promising for communications and teaching, though best practices and impactful uses are still evolving.</p>
<h4>Key takeaways:</h4>
<ul>
<li>AI tools can convert text to images and images to short videos.</li>
<li>Features include prompt enhancement and customization options.</li>
<li>Current limitations involve cost and API constraints.</li>
<li>Future use cases include education, marketing, and internal communication.</li>
<li>The technology is rapidly evolving with significant potential.</li>
</ul>

<p>The video presents the Smart Campus Team's summer projects, which leverage data sources such as TimeEdit, Eduroam connectivity, and SMHI weather data to enhance campus operations. By developing prototypes for forecasting occupancy, visualizing movement with heatmaps, and predicting energy use, the team highlights data-driven approaches to campus management. Additional tools introduced include navigation aids, virtual campus tours, streamlined service reporting, and AI-generated summaries to improve campus security.</p>
<p>The projects collectively illustrate how integrated data and new technologies can create a more efficient and responsive campus environment.</p>
<h4>Key takeaways:</h4>
<ul>
<li>Combining various data sources offers valuable campus insights.</li>
<li>Prototypes enable real-time occupancy and energy prediction.</li>
<li>New digital tools support navigation, reporting, and security.</li>
<li>The initiatives aim to boost campus efficiency and user experience.</li>
</ul>

<p>The video presents Digital Guardian, an AI-powered tool designed by Team Smart Campus to automate the analysis and summarization of campus security reports. By using OCR and advanced AI models, including GPT-4.1, it extracts key information—such as location, time, type, and severity of incidents—from uploaded PDFs, visualizes incident hotspots with heatmaps, and provides brief summaries and actionable recommendations.</p>
<p>The system currently runs locally to preserve sensitive data security. Future enhancements may include tighter integration with attendance and sensor data, automated feedback loops for improved AI accuracy, and streamlined notifications for urgent incidents. The project aims to reduce manual work and enhance pattern detection in security management.</p>
<h4>Key takeaways:</h4>
<ul>
<li>-AI streamlines security report analysis and visualization.</li>
<li>OCR and AI models extract and summarize key incidents.</li>
<li>Data privacy is maintained with local processing.</li>
<li>Integrating sensor and attendance data is a future goal.</li>
<li>Continuous feedback will improve AI performance.</li>
</ul>

<p>The team developed a prototype app that forecasts room availability by analyzing booking histories, WiFi (eduroam), and access data to estimate attendance and optimize space usage. These insights help improve scheduling and benefit students, staff, businesses, and future AI applications.</p>
<h4>Key takeaways:</h4>
<ul>
<li>Room usage data informs smarter space management.</li>
<li>The app predicts free times based on real usage patterns.</li>
<li>Attendance estimates combine WiFi and access data.</li>
<li>The system supports diverse campus and business needs.</li>
</ul>

<p>The video explains how Team Campus developed a heatmap tool using Eduroam Wi-Fi data to visualize campus presence and movement patterns. By mapping access points and using floor plans, they could see where and when students gather, such as during lunch breaks or in group rooms. The team combined this with booking data and CO2 levels to estimate which study spaces are available in real time.</p>
<p>They envision integrating these insights into a campus navigation app to help students efficiently find study spaces, quiet zones, and other facilities. The heatmap data also aids campus management in understanding building usage and planning for cleaning or renovations.</p>
<h4>Key takeaways:</h4>
<ul>
<li>Eduroam Wi-Fi data can help create detailed campus heatmaps.</li>
<li>CO2 sensors and booking systems aid in identifying available study spaces.</li>
<li>Visualizations support both student navigation and campus planning.</li>
<li>Future apps could make finding study or meeting spaces quicker and easier.</li>
</ul>

<p>This video explains how the Smart Campus team uses data-driven models, particularly a random forest algorithm, to predict campus room usage and movement flows. By integrating data from sources like CO2 sensors, Eduroam WiFi, and weather services (SMHI), the system forecasts occupancy, movement between buildings, and energy needs.</p>
<p>The approach allows for optimizing bookings, maintenance, accessibility, emergency planning, and even food service logistics. It also highlights challenges like data quality and privacy but emphasizes flexibility and a range of possible applications, such as guiding people to shorter lunch lines or preemptively adjusting ventilation. The main message is that leveraging diverse data sources can lead to smarter, more efficient, and adaptable campus environments.</p>
<h4>Key takeaways:</h4>
<ul>
<li>Predictive models use sensor, WiFi, and weather data to forecast campus usage.</li>
<li>Applications include energy optimization, booking, accessibility, and event planning.</li>
<li>Data quality and privacy remain challenges.</li>
<li>The flexible system can inform both real-time decisions and long-term planning.</li>
</ul>

<p>The video presents the Campus Restaurant Simulator, a tool designed to optimize restaurant operations on campus by leveraging various data sources such as SMHI, Eduroam, TimeEdit, and CO2 sensors. The simulator aims to predict customer flow and restaurant demand to reduce food waste, improve staffing efficiency, and enhance customer experience by providing estimated wait times.</p>
<p>Although the prototype currently uses synthetic data, the team plans to incorporate real data from multiple sources and address challenges like data quality and integration. The tool offers value to both restaurant operators and external stakeholders by enabling better resource planning and strategic menu decisions, and facilitates data-driven exchanges for mutual benefit.</p>
<h4>Key takeaways:</h4>
<ul>
<li>The simulator predicts restaurant demand using diverse data.</li>
<li>Main benefits include reduced waste, better staffing, and improved customer satisfaction.</li>
<li>Data quality and integration are ongoing challenges.</li>
<li>The approach provides commercial and operational value.</li>
<li>Future developments may include live wait-time estimates via cameras and other sensors.</li>
</ul>

<p>The video explains Team Smart Campus’s Energy Predictor, which uses machine learning (XGBoost) to forecast heating needs in campus buildings, aiming to optimize comfort and sustainability. The model analyzes historical energy, temperature, and simulated occupancy data to predict heat usage with high accuracy.</p>
<p>The team highlights the potential benefits of smarter, proactive heating and discusses expanding the model with real occupancy data, weather forecasts, and feedback systems to improve predictions and control. They emphasize the need for further collaboration and data integration for even better results.</p>
<h4>Key takeaways:</h4>
<ul>
<li>Machine learning can accurately predict building heat demand</li>
<li>Integrating weather and real occupancy data improves results</li>
<li>Automated, data-driven energy control aids comfort and sustainability</li>
<li>Human oversight and feedback systems remain important</li>
</ul>

<p>The Smart Campus team introduces a digital tourist guide designed for new Chalmers students in Gothenburg. The guide helps users find restaurants, events, and places around the city by integrating Google Maps and real-time event listings using APIs and web scrapers. Users can search, filter, and get directions for walking, driving, or using public transport. Future plans include personal recommendations, account registration, calendar integration, and social sharing features. The project aims to broaden student life beyond campus and make exploring Gothenburg easier and more engaging.</p>
<h4>Key takeaways:</h4>
<ul>
<li>Interactive city guide for Chalmers students</li>
<li>Integrates maps and event listings via APIs</li>
<li>Search, filter, and get directions for events and places</li>
<li>Plans for personalized recommendations and calendar features</li>
<li>Aims to connect campus and city experiences</li>
</ul>

<p>The video introduces the Virtual Campus project by Team Smart Campus, which allows users to explore the Chalmers campus virtually via computer or mobile. Viewers learn how prospective and current students can navigate campus buildings, ask questions about study programs, and receive detailed information about facilities and places of interest.</p>
<p>The platform uses 360-degree photos and interactive guides, with future plans to integrate more features such as custom tour routes, event info, and real-time data like restaurant queues. The project aims to enhance campus accessibility and engagement for both new and existing students, supporting use cases from digital open houses to finding amenities.</p>
<h4>Key takeaways:</h4>
<ul>
<li>Virtual Campus provides interactive, virtual exploration of Chalmers campus.</li>
<li>Users can receive personalized information and guided campus tours.</li>
<li>The project aims to combine various smart features for a comprehensive experience.</li>
<li>Future plans include event integration and real-time campus data.</li>
</ul>

<p>The video introduces Pathfinder, a project aimed at improving indoor campus navigation by creating a detailed graph-based network of rooms and connections, using floor plans processed by large language models. This system can generate optimal routes, update easily after renovations, and assess accessibility. Integrating positioning data from Wi-Fi access points and possibly 360-degree room photos could make location tracking more precise.</p>
<p>Pathfinder’s approach complements other projects like Chalmers GO, which focuses on route visualization. Together, these efforts aim to make campus navigation intuitive and accessible.</p>
<h4>Key takeaways:</h4>
<ul>
<li>Pathfinder uses graph networks for campus navigation.</li>
<li>Floor plans processed by AI generate room connection data.</li>
<li>Wi-Fi and 360 photos enhance location accuracy.</li>
<li>Projects aim for seamless, accessible indoor navigation.</li>
</ul>

<p>Campus Vision is an AI-powered tool developed by the Teams Smart Campus to enhance campus navigation and resource discovery. By processing 360-degree images with Google AI, it detects objects and provides detailed room information, such as equipment and available spaces. The system integrates data sources like Eduroam, CO2, and weather, and uses rule-based filtering, a blitz-net architecture, and Gemini for user queries.</p>
<p>Key challenges include manual image collection and distortion in panoramic photos, but future developments may automate this with robots or staff. Campus Vision supports digital twins, helps users find suitable rooms or inventory items, and can be integrated with booking and campus management tools.</p>
<h4>Key takeaways:</h4>
<ul>
<li>Uses AI and panoramic images for campus resource mapping</li>
<li>Integrates with multiple campus data sources for real-time info</li>
<li>Helps users search, view, and book suitable rooms and equipment</li>
<li>Plans exist to reduce manual data collection with automation</li>
</ul>

<p>This video introduces Chalmers Go, an app designed to help students and visitors navigate the Chalmers campus using indoor augmented reality (AR) directions through their phone's camera. It addresses the challenges of finding rooms, especially in complex buildings, and explains the technical aspects behind indoor navigation, such as computer vision, SIFT features, homography, and the use of Building Information Models (BIM).</p>
<p>The app aims to provide accessible routes for all users, integrate with occupancy data, and expand to features like guided tours and potential collaborations with other campus technologies like robots and drones. The main message emphasizes how combining these digital tools enhances the campus experience and connectivity.</p>
<h4>Key takeaways:</h4>
<ul>
<li>Chalmers Go offers indoor AR navigation for campus buildings.</li>
<li>Uses computer vision and BIM for accuracy.</li>
<li>Can adapt routes for various physical abilities.</li>
<li>Has potential integrations with other tech, like campus robots.</li>
<li>Enhances the digital and user experience on campus.</li>
</ul>

<p>The video introduces Ghostbuster, a chatbot developed for Chalmers to help users efficiently find the right contact person for various questions, such as academic advisors or finance contacts. It gathers and updates information from the university's web pages and personal profiles, aiming to improve on traditional website searches, which are often ineffective for specific inquiries.</p>
<p>Ghostbuster can display contact details and potentially handle administrative tasks or connect users to experts directly through chat. The team discusses ideas for future improvements, like automated data updates and integrating more personalized staff information.</p>
<h4>Key takeaways:</h4>
<ul>
<li>Ghostbuster helps users easily find relevant contacts at Chalmers.</li>
<li>It relies on web data and personal profile updates.</li>
<li>The chatbot improves on traditional search methods.</li>
<li>Future plans include automation and expanded functionality.</li>
</ul>

<p>The video introduces ServiceFlow, a proposed IT system upgrade for campus service management that leverages AI and better data collection. By automating data entry, tracking repairs, and analyzing fault reports, ServiceFlow streamlines task assignment, predicts common failures, and suggests solutions using historical data. AI-powered features include creating technician routes, prioritizing cases, sharing knowledge between senior and junior staff, and linking with existing tools like shared mailboxes and location services.</p>
<p>The system aims to boost efficiency, sustainability, and knowledge transfer while making reporting and resolving service issues simpler for everyone on campus.</p>
<h4>Key takeaways:</h4>
<ul>
<li>ServiceFlow collects and analyzes service data for better task management.</li>
<li>AI predicts issues and recommends solutions based on past cases.</li>
<li>Enhanced knowledge sharing between experienced and new technicians.</li>
<li>Integration with existing tools improves reporting and efficiency.</li>
</ul>

Team Smart Uni

Team 10xIndividual

Team Data Exploit

Team Education

Team Upskill

Team AI-based feedback

Explorative AI projects 2025

Choose subject

Course syllabus for

Subject

Each academic year contains different courses

See programme page at universityadmissions.se

Educational area

Permissible substitutions of compulsory courses

To see your programme syllabus – select the academic year in which you started the programme.

[en: page.studyprogrammepage.syllabusesview.coursesdescription]

Explorative AI projects 2025

Team Smart Campus

Project overview

Video description: Project overview

About Team Smart Campus

Video description: About Team Smart Campus

Digital guardian

Video description: Digital guardian

Room usage

Video description: Room usage

Heat map

Video description: Heat map

Campus Forecast and Flow

Video description: Campus forecast and flow

Restaurant-SIM

Video description: Restaurant-SIM

Energy predictor

Videobeskrivning: Energy predictor

Tourist guide

Video description: Tourist guide

Virtual campus

Video description: Virtual campus

Pathfinder

Video description: Pathfinder

Campus vision

Video description: Campus vision

Chalmers Go

Video description: Chalmers Go

Ghost buster

Video description: Ghost buster

Service flow

Video description: Service flow