Track C

I delve into the strategic use of historical accident data to identify high-risk zones. This session showcases how the Bangkok Metropolitan Administration has employed GIS analysis on historical accident data to pinpoint areas with a higher incidence of accidents. The insights derived from this analysis inform targeted physical improvements, contributing to a safer urban environment. Additionally, I investigate the power of analyzing retrospective traffic data to identify congestion hotspots. By examining historical traffic patterns, we pinpoint road segments experiencing the highest congestion. This granular understanding enables informed decision-making in urban planning and traffic management, offering practical solutions to alleviate congestion. These insights are derived from GIS and Map API.

HasiruAqua is a startup that is committed to providing next-generation technology to the agriculture and aquaculture ecosystem under low-cost or affordable business models that enrich the lives of farmers and consumers. Our two-fold mission is to improve the lives of farmers through economic transformation and to improve the quality of nutrition that citizens consume.

We’ve been working on revolutionizing the way of cultivating and consuming fish.Current practices of harvesting from natural sources are not only causing severe ecological damage but are also unsustainable and leading to reduced supply and increased pricing. HasiruAqua cultivates safe, hygienic fish by partnering with farmers and using technology. We offer farmers a “Full-Stack Fish Farming” service that helps them through the entire cultivation journey. We sell this produce to distributors, exporters and D2C brands at competitive rates and share profits with the farmers. Thus we are able to create an endless supply of farmed fish and help the farmer increase his annual income.

Integrating open mapping tools we have mapped 20k ponds which can help grow 20,000 tons of fish and other aquaculture products per year and increase the farmer income by 45% year on year. Combining Open Street Mapping tools and Hasiru Aqua technology data we can provide Anticipatory Action for Disaster Mitigation for farmers which will help reduce the risk of losing produce.

The “Battle of the Best OSM Mobile Data Collection Tool” is a workshop that aims to delve into two cutting-edge mobile data collection tools for OpenStreetMap: Organic Maps and EveryDoor. This workshop will serve as a platform for participants to gain insights, hands-on experience, and a comprehensive understanding of these innovative approaches to OSM data enrichment. By the end of the workshop, participants will not only grasp the technical intricacies but also understand the overarching societal and technological impacts of enhanced OSM data collection.

The workshop would feature the usage of two OSM mobile data collection tools, Organic Maps and EveryDoor. The functions of the said tools will be discussed then an actual utilization of them will be performed by the workshop participants. The participants are expected to share their experiences in testing the tools.

In geospatial analysis, OpenStreetMap (OSM) data have become essential for urban planning, environmental assessment, and location-based services. However, these tasks often require a complex interplay between human expertise and technical tools. To bridge this gap, we present OSM-GPT, a groundbreaking opensource tool that seamlessly integrates artificial intelligence, particularly the ChatGPT API, into the geospatial domain. This fusion enables users to effortlessly generate OSM queries, extract OSM data, and visualize it in a custom and interactive manner.

## Key Features:

**AI-Powered Query Generation:**
OSM-GPT’s hallmark feature is its ability to generate OSM queries via the ChatGPT API. Users can simply describe their requirements, such as “get all buildings” or “extract road networks”, and OSM-GPT will transform these descriptions into actionable queries.

**Effortless Data Extraction:**
Overpass API is used to retrieve OSM data using the queries generated by OSM-GPT. Each query generates a dedicated layer, simplifying data management and enhancing the visualization experience.

**Manual Query Support:**
While AI-generated queries are a highlight, OSM-GPT also caters to users familiar with OSM query syntax. This enables the manual input of queries for extracting specific OSM data, ensuring flexibility and catering to a wide range of user expertise levels.

**Custom Layer Creation:**
To offer a comprehensive view of the data, OSM-GPT empowers users to customize layer colors. This characteristic improves the distinction of data and assists in effectively conveying valuable insights.

**Selective GeoJSON Downloads:**
OSM-GPT enhances data utilization by allowing users to download GeoJSON files tailored to their needs. Users can choose to download only polygons, lines, points, or all features.

## Future Prospects
OSM-GPT’s roadmap includes plans to introduce custom layer overlay functionality. The integration of users’ own spatial data with OSM extracts will facilitate dynamic and insightful visualizations.

## Use Case Scenario:
In times of disaster, OSM-GPT can be a lifesaver. It quickly figures out where buildings, roads, and hospitals are in the affected area without volunteers spending time searching. These key details are then displayed on a map, helping responders decide where to send help and where there are clear roads. Volunteers don’t need to worry about complicated queries; they can focus on taking action.

**In conclusion**,
OSM-GPT marks a groundbreaking advancement in making geospatial discovery and analysis accessible to a wider audience. By integrating AI capabilities with OSM data manipulation, the tool streamlines the process of query generation, data extraction, visualization, and customization. It empowers users of varying technical backgrounds to harness the power of geospatial data and transform it into actionable insights.

Demo: [https://dub.sh/osm-gpt-demo](https://dub.sh/osm-gpt-demo)

Source Code: [https://github.com/rowheat02/osm-gpt](https://github.com/rowheat02/osm-gpt)

Try OSM-GPT: [https://dub.sh/try-osm-gpt](https://dub.sh/try-osm-gpt)

This research presents a comprehensive analysis of coastlines using OpenStreetMap (OSM) data and Sentinel satellite imagery. The research uses these two distinct yet complementary data sources to improve understanding and decision-making in coastal management and monitoring. The comparison of coastlines is a critical endeavor in geospatial analysis, essential for evaluating changes caused by natural processes and anthropogenic influences. The research investigates discrepancies and congruencies in coastline delineations using the rich dataset of OSM, a collaborative mapping platform, and high-resolution multispectral imagery from Sentinel satellites.
Through a systematic assessment, the study underscores the strengths and limitations of each data source. OSM, which is powered by volunteer contributions, provides fine-grained details but is influenced by data availability and update frequency. Sentinel imagery, on the other hand, provides a comprehensive view of the coastal landscape but necessitates specialized processing for accurate coastline extraction.
Furthermore, the research introduces the use of OpenStreetMap data as a reference dataset for validating the accuracy of Sentinel-derived coastlines. Using spatial analysis techniques, the study quantifies deviations and assesses the overall precision of Sentinel-derived coastlines. Normalized Difference Water Index (NDWI) is used to derive the coastline of the Capital City of Indonesia, Jakarta. There are four years data starting 2017 to 2020.
The findings of this research contribute to a more informed approach to coastline analysis by combining the crowd-sourced nature of OSM and the technological prowess of Sentinel satellites. The findings have implications for coastal management, disaster preparedness, and environmental assessment, guiding policy and planning efforts.
Finally, this study provides a thorough comparison and evaluation of coastlines using OpenStreetMap data and Sentinel imagery. The combination of these data sources highlights the importance of collaborative mapping and remote sensing in furthering our understanding of dynamic coastal environments.
Keywords: Coastline comparison, OpenStreetMap, Sentinel imagery, geospatial analysis, remote sensing, coastal management, environmental assessment, spatial analysis.

As traditional fossil fuels become increasingly scarce, the prioritization of renewable energy sources, like solar power, has gained significant importance in today’s global landscape. This research introduces a novel method for simulating solar radiation across rooftops at Paschimanchal Campus, utilizing Photogrammetry and Geographic Information System (GIS) techniques to estimate solar photovoltaic (PV) potential.
The approach employs an object-based technique to accurately extract and categorize buildings, leveraging high-resolution data such as orthophotos, UAV-derived Digital Surface Models, and Digital Terrain Models. To ensure precise projections of solar panel efficacy, it is crucial to evaluate solar radiation, with the conversion of this radiation into solar PV potential contingent on the panel’s technical attributes, such as whether it employs monocrystalline or polycrystalline technology.
Consequently, a comparative analysis is conducted, pitting the calculated expense of solar panel implementation against the cost of grid-based electricity within the study area. The findings reveal that among the buildings’ rooftops, 18 locations are exceptionally suitable for solar panel deployment, offering a combined usable area of 4303.67 square meters. In terms of power generation, the study area is projected to produce 868 MWh/yr and 1274 MWh/yr using polycrystalline and monocrystalline technologies, respectively.
Remarkably, just 1428.42 square meters of the identified suitable area would suffice to meet Paschimanchal Campus’s annual energy requirements. In the case of polycrystalline technology, approximately 876 panels are needed, with an associated cost of NRS.1,57,68,000. Through a cost analysis spanning a decade, the study concludes that solar energy is poised to outperform grid-based electricity by NRS.1,04,10,785, thereby emerging as a more cost-effective and sustainable energy alternative.