Eco-Intel: Harnessing Technology for Wildlife Conservation in East Africa
From the savannahs of Maasai Mara to the coral reefs of the Indian Ocean, East Africa is home to some of the most diverse and remarkable ecosystems. These landscapes provide habitats for terrestrial and marine wildlife which play a crucial role in global biodiversity. However, this natural wealth is confronted with challenges stemming from habitat destruction, climate change, poaching, and illegal fishing which are intensified by rapid population growth and urbanisation. These challenges could result in irreparable damage to East Africa’s wildlife and ecosystems, threatening both local communities and the broader ecological balance.
The urgency to address these challenges has pushed conservation efforts toward leveraging technology to design innovative strategies. This piece discusses how innovations like satellite remote sensing, GPS tracking, and internet of things (IoT) networks are revolutionising how conservationists monitor and protect ecosystems while offering new solutions for sustainability in East Africa. It also emphasises the importance of data quality, which is enhanced via data engineering, an important pre-requisite to the end use AI models.
Remote Sensing Technologies and Autonomous Vehicles
Remote sensing technologies play a crucial role in environmental monitoring and conservation efforts across East Africa which is essential for timely response to wildlife distress. A prime example is Digital Earth Africa, which demonstrates how earth observation (EO) data was employed to address challenges faced by communities and wildlife around Lake Baringo in Kenya. Rising lake levels began submerging an island towards the centre of the Lake which threatened nine Rothschild giraffes that lived on it. For communities, the rising lake levels meant some were unable to take their livestock to graze in familiar areas due to less land available. By leveraging satellite imagery and geospatial data, local authorities were able to monitor the changing landscape and relocate the giraffes to safer areas as well as guide the local community to other grazing areas, which would also reduce conflict with the newly relocated giraffes. This use of EO real-time data highlights how remote sensing can guide swift, effective decision-making in conservation efforts, particularly in response to climate-related challenges.
A similar initiative is the Global Monitoring for Environment and Security (GMES) and Africa Program which provides access to satellite-derived data for faster mitigation of environmental threats. The program supports fishery management in Kenya and Tanzania by providing ocean data to track aquatic populations, water temperatures, and environmental conditions. This initiative could be supplemented by drone and underwater autonomous vehicles (UAVs) to offer high-resolution and real-time observations for localised analysis of environmental changes. Fusing detailed snapshots of small regions from drone-collected data with lower resolution images covering large areas enables data engineers to create more robust datasets. This could benefit the region as it faces a substantial lack of quality data across various sectors.
Data engineering plays a crucial role in this process by ensuring that the raw data is cleaned, standardised, and integrated for use in advanced machine learning for further analysis. A high-level view of how the data engineering process would take place is outlined below:
1. Data pre-processing:
Cleaning and filtering: raw data often contains ‘noise’ (irrelevant or erroneous data) so algorithms are applied to filter out anomalies, ensuring the data is clean and reliable for analysis.
Georeferencing: involves drone mapping of data points to real-world coordinates. This ensures high-resolution images of a location to match exactly with satellite data from the same area.
Temporal synchronisation: differences in data collection times (satellite data is more periodic than drone data) are aligned to ensure both datasets represent the same time frames when combined.
2. Data integration
Multisource fusion: involves merging different layers of information to create a single, unified dataset reflecting both the broad patterns observed by satellites and finer details captured by drones.
Enhancing resolution: drone data is used to enhance lower-resolution satellite data in a process called ‘super-resolution.’ This improves overall dataset quality by providing more detail where satellite data is too coarse.
Handling scale differences: specialised algorithms are used to ensure the scales of datasets match correctly since satellite data covers large areas and drone data focuses on smaller regions. This allows for meaningful comparisons and analyses.
3. Data quality assurance
Validation and verification: measures such as cross-validation are employed to confirm that observations between drone and satellite data are consistent.
Error correction: discrepancies between data sources are resolved by algorithms.
Standardisation: applying common formats and units across drone and satellite data. This allows easy integration of additional data sources and sharing of datasets with researchers and institutions.
Following this, a pivotal strategy in mapping and analysing spatial data to inform conservation and environmental management is the use of geographic information systems (GIS). Modern GIS platforms offer tools that enable users to overlay, analyse, and visualise geospatial data with increasing precision.
Figure 1. Diagram showing how GIS platforms overlay different data areas.
The key to successful GIS implementation in the region lies in finding a platform that suits both the financial and technical capabilities of local conservation organisations. For a lower income region like East Africa, the most suitable GIS solutions should be able to balance cost and be easy to implement. Open-source platforms like Quantum Geographic Information System (QGIS) offer cost-effective alternatives due to free downloads and modifications. For example, custom mobile extensions like QField or Input for QGIS could be developed specifically for East African conservation contexts, enabling rangers and researchers to collect field data on wildlife sightings, poaching incidents, or environmental conditions. These mobile extensions could sync directly with the main QGIS platform, allowing data collected in the field to be immediately integrated into centralised databases. It supports a wide range of data formats that integrate with other open-source tools, thus eliminating the need for expensive software. This is in comparison to commercial systems like Environmental Systems Research Institute, Inc. (ESRI's) ArcGIS which provide enhanced features like premium high-resolution imagery and real-time data processing.
Understanding insights derived from this data is crucial for decision makers across various sectors. For example, by analysing patterns in wildlife movements or environmental changes, government agencies can develop more effective regulations and conservation strategies to preserve biodiversity. Local communities and stakeholders can optimise their resource management practices, helping to prevent overexploitation of vulnerable ecosystems. This data-driven approach ensures that both ecological sustainability and local livelihoods are maintained.
Tracking and Connectivity Solutions
In addition to drone usage, GPS tracking is a powerful tool allowing researchers to gather data on animal behaviour, migration routes, habitat usage, and responses to environmental changes. This data can provide insights into creating targeted conservation strategies such as by tracking endangered species to better design protected corridors for safe migration across fragmented habitats or identifying areas where human-wildlife conflict is likely to occur.
Following data collection for such initiatives, integration and processing using cloud-based platforms and distributed computing frameworks, like Hadoop or Apache Spark, helps to manage large geospatial datasets efficiently for analytics. Hadoop allows vast amounts of data to be stored and analysed across multiple machines. Apache Spark offers fast, in-memory data processing, making it ideal for real-time analytics and large-scale machine learning tasks. These tools ensure that as more data is collected, systems can scale proportionately to manage quantities which allows tracking efforts to cover more species and areas.
IoT networks are also proving to be transformative for conservation. A pioneering example is Akagera National Park in Rwanda, the world’s first ‘smart park’ powered by LoRaWAN technology. This IoT network provides real-time monitoring of wildlife, park rangers, and vehicles. Similarly, Northern Rangelands Trust in Kenya operates Africa’s largest IoT conservation network aiding in effective management of protected areas and monitoring threatened species across landscapes. By connecting sensors, cameras, and GPS collars through IoT technology, conservationists are better equipped to respond to threats like poaching and habitat destruction. The success of IoT networks hinges upon standardisation of data formats which ensure that data from devices and platforms can be seamlessly integrated. By adopting interoperability frameworks such as Open Geospatial Consortium (OGC) standards, conservation efforts benefit from enhanced data sharing, fostering collaboration across organisations and even countries. Implementing these best practices improves data quality, facilitating stronger regional cooperation and more effective conservation outcomes.
Conclusion
In the face of pressing conservational challenges, technological innovations such as remote sensing, GPS tracking, drones, and IoT networks are driving significant advancements towards conservation across East Africa. These tools enable researchers and conservationists to monitor ecosystems in real time, respond to environmental challenges efficiently, and develop data-driven strategies for protecting biodiversity. Continued investment in these technologies is essential for scaling their impact and ensuring sustainability in the face of ongoing environmental conflict. By adopting these innovations, East Africa can better safeguard its rich natural heritage for future generations.
