Highlighting use cases in which AI is powering drones to make an impact in biodiversity.
Seemingly overnight, AI is everywhere. In the past year, its influence has become more pronounced in our society. We see AI across social media platforms, on search engines, as new features in productivity apps such as Slack, Microsoft office, and Zoom. How did AI get here? Let’s dive into its beginnings.
The origins of AI date back to the 1950s and Alan Turing, a British mathematician who envisioned the possibility of artificial intelligence. In his paper, “Computing Machinery and Intelligence,” Turing asked the question, “Can machines think?” He theorized that machines could expand their capabilities beyond their original programming and developed the Turing test to assess whether a machine could simulate human intelligence. If a human observer could not discern whether responses came from a machine, then it was deemed intelligent. Unfortunately, Turing did not have access to the computing power necessary to test his theory.
These are the events that charted a path to the AI revolution of today:
Sentiments about AI
Given AI’s long history, I’m often surprised by the growing negativity towards its increased capabilities and advancements. According to Stanford’s Institute for Human-Centered Artificial Intelligence 2024 AI Index Report, “people across the globe are more cognizant of AI’s potential impact—and more nervous.” The report highlights an Ipsos survey where 52% expressed nervousness toward AI products and services, compared to 39% in 2022. Pew data suggests that 52% of Americans report feeling more concerned than excited about AI, rising from 38% in 2022.
The most cited concerns I hear stem from the fear that AI will take away jobs and the lack of standardization in AI governance. Despite these growing pains, there is a lot of social good that is coming from AI. Let’s review use cases where AI is augmenting the impact of drones in improving biodiversity efforts.
Applications of AI in Drones for Conservation and Disaster Response
Applications of AI in drones started in 2013 and focused on autonomous navigation and collision avoidance. As the capabilities of AI and drones have progressed, researchers and experts have developed new approaches for combining drone technology with artificial intelligence to transform efforts in wildlife conservation and natural disaster recovery.
Use Case #1: Wildlife Monitoring
T. Petso and R. S. Jamisola from the Botswana International University of Science and Technology developed an AI-powered approach to identify wildlife herds based on unique behavioral patterns, as reported in Science Robotics. This system works by assigning point values to each animal; it combines this information along with pattern analysis of movements, interactions, and behaviors to identify and classify animal species. This analysis relies on video footage collected from an aerial view and is made possible by drone cameras.
The benefit of using this approach is that it provides a more reliable alternative for population counts and facilitates conservation planning. Existing methods for identifying individual animals face accuracy challenges due to issues like natural camouflage, foliage obstruction, and limitations in drone height. The AI is using additional data points to make a more accurate determination of the species captured by drone footage. At Botswana’s Khama Rhino Sanctuary, this approach has been successfully validated, enabling researchers to classify various species including zebras, giraffes, and elephants. Since this AI model identifies animals based on movement patterns, it can also be applied to classify human activities, leading us to our second use case.
Use Case #2: Counter Poaching
Traditionally, anti-poaching efforts have depended on ranger foot patrols and manual surveillance monitoring. These methods have coverage limitations and rely on the number of personnel and their stamina. T. Petso and R. S. Jamisola’s AI powered approach to identify herds of wildlife has applications to combat poaching activity. The AI can differentiate between patrolling rangers and potential poachers by analyzing movement patterns. The AI’s ability to identify poaching activity can be used to amplify existing anti-poaching strategies.
- Predict poaching activity: The AI could be further developed to integrate with tools such as PAWS (Protection Assistant for Wildlife Safety) which works with the open-source SMART — or Spatial Monitoring and Reporting Tool. This integration could further refine how patrol teams schedule their patrol routes and predict future poaching threats. It can also be used to advocate for stronger support from local communities, law enforcement, and government agencies.
- Thermal Imaging Integration: The original study utilized visible light cameras, but an application where thermal imaging cameras are combined with drones and AI could lead to an autonomous system that could be used around the clock, especially at times when poaching activity is the highest.
- Multi-point systems: Another interesting application would be to explore integrating aerial and ground-based camera platforms, such as TrailGuard or Hack the Planet’s AI camera trap, into a single system that could serve to achieve multiple conversation efforts. The AI could be trained to ingest data from ground-based systems and provide a more holistic view of poaching activity and further refine predictive capabilities. Additional surveillance forms can help overcome geographic challenges.
In order to implement these use cases, the following challenges would need to be addressed:
- Ensuring adequate network bandwidth and reliable communication channels.
- Designing drones that meet specific size requirements to minimize wildlife disturbance.
- Amending local legislation and adjusting permit protocols to support drone usage in conservation efforts.
Use Case #3: Leveraging AI to Enhance Disaster Response Efficiency
Our last use case highlights how AI can analyze drone footage to guide first responders with deploying resources more effectively. Drones provide critical insights by assessing damage, mapping escape routes, directing rescue efforts, and locating survivors. AI models process this footage to predict outcomes and support decision-making, such as advising on the distribution of food, medical supplies, and personnel based on the severity of impact in different areas. This capability is especially valuable when emergency teams are handling multiple simultaneous incidents, allowing them to allocate resources efficiently across various disaster zones.
AI can also analyze data following a natural disaster to better comprehend what happened, identify vulnerabilities, and help prepare for future incidents. The potential for AI extends beyond analysis and prediction to include scenario planning, and there is room to further integrate AI with other robotic technologies.
These applications showcase several ways AI is enhancing the capabilities of drones. AI serves as a powerful accelerator that can amplify biodiversity initiatives. As AI continues to advance, our main challenge and responsibility is to ensure that systems, processes, and tools proactively adapt to accommodate these innovations.
