Introduction:

Access to accurate and navigable road network data is crucial for both personal transportation and corporate logistics, as well as for efficient humanitarian and relief efforts. META (previously known as FACEBOOK) has been contributing to the mapping and technology ecosystem, improving the quality and coverage of data in OpenStreetMap and supporting the development of Free Software tools. The AI-Assisted Road project introduced by META in 2019 allows the community to make use of predicted roads from satellite imagery, facilitating road mapping in OpenStreetMap. Skilled mappers from OpenMap Development Tanzania (OMDTZ) have been consistently working on digitization and validation of the Tanzania road network data, leading to successful implementation of various projects in the region.

Tools and Methodology:

• MapwithAI Tasking Manager: Used for mapping projects.

• Rapid editor: An AI-assisted road mapping tool available in the tasking manager.

• JOSM: An advanced OpenStreetMap editor for data validation.

• OSMAnd: A map and navigation application for Android and iOS.

• Open Data Kit (ODK): A free application used for data collection, combined with Kobo toolbox servers and ODK central servers.

Mapping initiatives carried out in Tanzania:

1. Crop Mapping Project:
   Funded by GEOGLAM, this project involves identifying crops in selected areas through field validation. The collected data is used for monitoring and improving crop forecasting models and agricultural assessments, as well as promoting modern agricultural techniques and adaptation to climate change.

2. Solid Waste Management/Litter Mapping Projects:
   Roads were digitized and updated to facilitate surveys and assessments of solid waste practices. Roads were included in the basemap along with buildings for easier navigation to areas of interest.

3. School Feeding Programme Mapping:
   A project financed by WFP to capture food program practices in pre and primary schools. Existing road information in OpenStreetMap was used to reach school locations, and data was collected using ODK and OSMAnd.

4. Mills Census Study:
   A nationwide survey to obtain the spatial distribution of mill machines. FOSS tools like ODK and OSMAnd were used for data collection, and QGIS for data cleaning and analysis. Road networks supported navigation and travel cost estimation.

Contributions of OpenStreetMap roads to project success:

• Locality navigation.

• Estimation of transport costs.

• Prediction of the mode of transport.

• Calculation of travel time, reducing costs and increasing comfort.

Challenges and Solutions:
- Quality checking controls: Errors in road data required mass validation by skilled mappers and validators.

• Imagery clarity: Blurry Maxar imagery made it difficult to distinguish between highways and waterways in some projects.

• Data conflicts with other OSM users: Flagging issues to Meta tech experts helped resolve conflicts.

• Remoteness of some areas: Poorly constructed roads in rural areas hindered data collection.

• Unpredictable climatic weather conditions: Unfamiliar climate and unpredictable rain seasons affected data collection.

• Presence of wild animals: Precautions needed in areas with wildlife, requiring additional details for road connectivity and navigation.