Software, data and AI that work every day.

Location Science is a software and data company from Upper Bavaria. We digitalise processes, introduce AI and make geodata usable, for public authorities and for companies. With StoryMapp we have a product of our own that shows how we work.

Does this sound familiar?

That is exactly where we start: with interfaces, automation and software that fits your task.

Open data shows the ways. Context decides which one is right.

3D terrain map with a marked tour route over mountains and through a town, rendered by us from open data.
Rendered by us from open data: terrain model and trails from OpenStreetMap, the places from our product StoryMapp, the route from our tour planner.

What we are working on

Four projects that show how we bring data, software and AI together. Two of them are still in development, and we say so.

  • Tour planner app: a planned 16.6 km hiking route on a map, trail sections coloured by surface, stops and a waypoint marked.
    In development

    Tour planner that builds the tour for the day

    Most tour portals show a database of ready-made tours. Our planner creates the tour for the user and the day. It combines weather, warnings and closures with fitness, time budget and preferences, and with trail surface, gradient, shade and places to rest or eat.

    The result is a route that fits, instead of a list to search through.

  • Mobile 3D ski-area map with pistes, lifts and video points along the run.
    In development

    SkiApp: the ski day, measured and shared

    The app records the skier's statistics and health data and links them to the geography of the ski area. Photos and videos are embedded in a 3D replay of the run that can be shared with friends and followers.

    Behind it: a terrain model, piste data, on-device positioning, and a snow and weather report from several sources.

  • StoryMapp app: map of Upper Bavaria with round photo markers for places with stories.
    In operation

    StoryMapp: stories about places, worldwide

    Our own product, free for iOS and Android. More than 8 million stories in six languages, worldwide coverage, a monthly update cycle, users in 58 countries. Partner: the artists' project Art Hop Silver Lake in Los Angeles.

    The data pipeline behind it runs without manual work, and the users' location never leaves the device. About StoryMapp →

  • In development

    Emergency map for disaster response

    A map that works without network and without power, and is still current. It gives emergency services orientation and a basis for planning during a large-scale blackout, on the device, from current open data.

Three founders. Engineers first.

We come from geography and geoinformatics, cartography and information and communication technology. We have worked in start-ups as well as in large companies, often on systems where mistakes are expensive.

  • Geography and geoinformatics
  • Cartography
  • Information and communication technology

Working experience: public-safety organisations (BOS), defence engineering, global car navigation, DevOps.

Tobias Neumann

Managing partner

Studied geography (B.Sc.) and geoinformatics (M.Sc.) and has worked with spatial information ever since. Experience from equipping emergency control centres, developing air-defence systems and global car-navigation projects, including for TomTom. Most recently research on open geodata at a world-leading technology company. Builds and optimises spatial databases, geodata infrastructures and data compilers that turn raw data into deliverable map formats, as well as front ends for map display. A particular focus is proofs of concept that make new ideas quick to test, and agentic systems for autonomous operation.

Kangle Li, PhD

Founder

PhD in information and communication technology, with research focused on designing and implementing digital signal processing algorithms. Research at several universities, including Munich University of Applied Sciences (Hochschule München). In-depth expertise in pattern recognition, machine learning and deep learning, applied to object detection, classification and natural language processing. Builds applications on large language models end to end, from vector databases, frameworks and model APIs to deployment and fine-tuning. A particular focus is the control layer around the model that reduces hallucinations and keeps output quality reliable.

Working with us

Three ways to work with us. Each begins with an email, and one of the founders always answers.

  • Daily rate €900 net, for development and consulting alike.
  • Fixed price for clearly specified projects under a works contract, based on a jointly written specification.
  • First conversation free of charge, remote or at your office.