Simple web map application showning information about main buildings of the University of Pardubice campus for first time visitors. Important off-campus buildings in Pardubice are included as well. On a location-aware device it can show your own location on the map and can be used for fínding the way in the campus. The app contains short descriptions of the buildings with links for more information on University website, and the user can extend/update this information using Gemini AI.
The screenshot shows information pop-up of a faculty. The text next to the blue vertical line is AI generated. The text obove it and the link below is static.
- Current development version has the basic functionality completed.
- Known issues: the AI model likes to hallucinate links, so links in the AI generated part are often invalid. Other information like e-mails, transport information, etc., seem to be correct. The current position does not show on my mobile phone.
- Planned feature - ability to ask additional questions related to the opened map point with a prompt.
UPCE_campus.html - Place this file in a place accessible to users, users open the file in a web browser. It is stand-alone html app, no other files required. Every user has his own updated/extended vesrsions of building descriptions in his browser storage. The html file itself can be stored anywhere (i.e. local/network disk, user's phone storage, or published on a web server).
Moving around:
The map starts zoomed to the campus. Except common mouse or touch gestures you can use the buttons on the right, the home icon zooms on the campus, the globe icon to zooms out to see all POIs.
AI text:
Clicking on a named POI you get information pop-up. The pop-up contains short decription of the Universiy facility you clicked, link to its main web page, and button to generate additional information using AI. To employ this functionality, you have to first set API key, see next chapter.
Language and Gemini API Key
If you want to generete updated/extended building descriptions, you need to use your own Gemini API Key. If you do not have any, generate it in Google AI Studio and insert it in the when asked. The process involves creating a project (simply create in Projects and give it a name) and then create the key for that project in API Keys. Copy the text string of the key to the Gemini API Key field in the app when asked. Free api key is enough for using the app (20 AI requests/day). More detailed documentaion. Settings through the gear icon on top right have links to do it in-app.
The other button with language name switches to the language on the button label, supported languages are Czech and English, English is default.