Oliver Förster · Geoinformatics · 3D geospatial data · Geospatial services

Geospatial data analysis, reproducible processing and service-based delivery.

I design and implement custom solutions for spatial data. My work focuses on semantic 3D city models, CityGML, Python-based GIS methods, and standardised or application-specific web services.

Illustration of a Linux penguin in front of a cityscape with a laptop displaying the Blender interface
Illustrative subject visualisation · Linux · Blender · 3D GIS

Professional focus

Domain analysis, software engineering and system operations

My work combines the domain-specific analysis of spatial data with technical implementation. The data model, geometry, coordinate reference system, semantics, interfaces and operating environment are considered as a coherent system.

For tasks that cannot be addressed adequately with existing standard tools, I develop specialised Python, GIS and 3D applications. This includes data conversion, validation, spatial analysis, geometric processing, texturing and automated quality assurance.

Experience in Linux, web and systems administration enables me to package, secure, monitor and deploy processing methods in a defined and reproducible operating environment.

Areas of work

Aligning the data model, processing methods and system operations

Technical solutions are derived from the domain data model, intended use cases and verifiable quality requirements.

01 · Data and models

Assess geometry and semantics

Analysis of geometries, semantic structures, coordinate reference systems and exchange formats, particularly for CityGML, CityJSON, point clouds, raster data and PostGIS.

02 · Methods and software

Implement reproducible processing

Development of reproducible workflows, Python applications and extensions for QGIS and Blender, including validation, error handling and automated testing.

03 · Services and operations

Provide controlled interfaces

Delivery as a web application, programmatic interface or WPS, taking account of access control, logging, resource limits and monitoring.

Web Processing Service

Custom geoprocessing through standardised interfaces

I develop custom WPS processes and Python-based geoprocessing services for recurring or computationally intensive tasks. Inputs, parameters and domain constraints are validated explicitly; processing and output follow defined, reproducible rules.

Depending on the task, I combine PyWPS, QGIS Server, QWC, PostgreSQL/PostGIS and standalone Python modules. The technical implementation includes authentication, logging, resource limits, temporary storage, cleanup and monitoring.

Interface
WPS based on the OGC specification or an application-specific geoprocessing API
Processing
Python, QGIS, PostGIS and command-line methods with defined inputs and outputs
Validation
Formats, geometries, coordinate reference systems, file sizes and domain constraints
Operations
Authentication, logging, resource limits, temporary storage, cleanup and monitoring

3D geospatial data

Systematic processing of semantic 3D city models

The workflow is defined as a continuous processing chain from initial assessment to delivery. Domain and technical checks are applied at every processing stage.

01

Assessment

Review of the data model, level of detail, semantics, georeferencing and geometric consistency.

02

Transformation

Import, conversion, modelling, texturing and derivation of supplementary geometry.

03

Quality assurance

Validation, error analysis, defined fallback procedures and traceable processing.

04

Delivery

Export to standardised formats and integration into databases, web services and visualisations.

Technical focus

Selecting technologies according to domain and operational requirements

Open-source components and open standards are preferred. Existing proprietary environments are integrated where required by the data flow, interfaces or system landscape.

PythonQGISQGIS ServerQWCPyWPS / WPSPostgreSQL / PostGISBlenderCityGML 2.0 / 3.0CityJSON3DCityDBLAS / LAZDTM / DSMFMEGeoServerOpenLayers / CesiumJSLinux / Docker / ApacheMachine learning / computer vision

Working method

Assessing requirements, test datasets and results transparently

Technical decisions are based on explicit constraints, representative data and reproducible tests.

Structured analysis

Complex tasks are divided into verifiable subproblems, data flows and interfaces.

Testing with representative data

Prototypes are evaluated early using representative datasets and known edge cases.

Measurable optimisation

Data quality, runtime, memory use and error tolerance are improved against defined criteria.

Reproducible operations

Configuration, dependencies, logging and reproducibility are part of the delivered solution.

Contact

Technical discussion and project enquiries

I am available by email for enquiries concerning geodata, 3D GIS, CityGML, automation or Web Processing Services.

Send an email

Enlarged project view