Imagine a future where Public Safety officers have a powerful new tool at their disposal: a smart UAV, or unmanned aerial vehicle, that can be effortlessly controlled with voice commands. This is the vision behind a research project funded by CCI Central Virginia Node, and it holds immense potential for revolutionizing how Public Safety operates in dynamic urban environments.
Kostadin Damevski, Ph.D. an associate professor in the department of computer science at Virginia Commonwealth University School of Engineering, leads the project. His expertise in software maintenance and evolution is crucial for ensuring the system’s long-term success.
The project tackles the challenges of integrating UAVs into the daily routines of Public Safety officers. Traditionally, operating a drone requires a dedicated pilot, hindering its practicality in fast-paced situations. Dr. Damevski’s research aims to overcome this limitation by developing a system that allows officers to control the UAV using high-level voice commands.
The core functionality revolves around two key aspects:
- Intuitive Voice Control: Public Safety officers on the ground will be able to issue commands to the UAV using natural language. Imagine saying “monitor that building” or “explore the building’s edge” to direct the UAV’s flight path. This eliminates the need for complex joysticks or controllers, making UAV operation seamless and intuitive.
- Autonomous Navigation: The UAV won’t be a passive tool; it will be equipped with autonomous navigation capabilities. This means the UAV can navigate its surroundings without constant human intervention, avoiding obstacles and maneuvering safely through urban environments.
While the project is still in its early stages, the one-year mark has yielded promising results. The system can now recognize basic commands related to a specific building, such as monitoring it, exploring its edges, or capturing pictures. To achieve this, the researchers leverage an off-the-shelf toolkit called Raza, specifically designed for voice command recognition. Additionally, they incorporate OpenAI’s powerful generative pre-trained model GPT-3 to further enhance the accuracy of command recognition.
Looking ahead, the project delves into even more exciting possibilities. The future roadmap includes exploring voice-based programming of UAVs, enabling officers to provide more complex instructions. Additionally, the researchers envision scenarios involving multiple UAVs and human operators working together on reconnaissance missions.
The successful integration of voice-controlled, autonomous UAVs has the potential to significantly enhance public safety operations. Imagine SAR missions where UAVs can quickly locate missing persons in complex urban areas. Envision faster response times to emergencies as UAVs provide aerial surveillance and relay critical information to officers on the ground. The possibilities are vast, and this research project, led by Dr. Damevski’s expertise, paves the way for a future where smart UAVs become a vital asset in the public safety officer’s toolkit.
Watch Kostadin Damevski Ph. D.’s talk on smart UAVs for public safety here.
