Jump To Key Section
- More About Dmytro and the path that led him to develop an interest in AI and biodefense
- What motivated Dmytro to focus specifically on biodefense?
- Dymtro’s proposed AI-based biological detection and defense software system
- How is it different from the current public health monitoring systems?
- Major hurdles faced in making this vision a reality
Among recent interactions, it was a delightful brainstorming session with Dmytro Zaharnytskyi, an AI and machine learning engineer with a forward-thinking vision for the detection and biodefense areas. Dmytro, with a versatile background in AI solutions for public health and military use, aims to strengthen his efforts toward developing AI-based software for biological detection and defense systems. The discussion included his past experiences, motivation, and future aspirations.
More About Dmytro and the path that led him to develop an interest in AI and biodefense
His road into technology commenced with a bachelor’s degree in cybersecurity, imparting a robust foundation concerning challenges in securing critical infrastructure. Then, he earned a master’s in AI systems, which opened my horizon towards machine learning and its applications. For the last five years, he has been involved in a large number of sets concerning machine learning operations, biotechnology, and anomaly detection, particularly focusing on military safety and public health. He saw AI in the form of computer vision models for object detection on the battlefield, and that became his eureka moment, seeing how AI could save lives in risky situations.
What motivated Dmytro to focus specifically on biodefense?
The passion is motivated by the recognition that biological threats pose a terribly great challenge to modern society – a pandemic or bioterrorism. The COVID-19 pandemic was a vivid demonstration of how quickly a biological threat can overload health systems in different parts of the world. His goal is to develop an AI-based software system that performs early threat detection by analyzing anomalies in environmental and hospital data. Upon identifying those anomalies, a faster response could be initiated that could benefit thousands, perhaps millions, of lives.
Dymtro’s proposed AI-based biological detection and defense software system
The idea is to create a sophisticated detection and response paradigm using state-of-the-art AI technologies for identifying anomalies. The system will include environmental sensors and be able to analyze real-time data streams from hospitals. For example, if a hospital network were to report unusual symptoms or show a certain environmental sensor to be indicative of an unanticipated chemical marker, then the AI would pull those dots together to potentially identify a threat.
Must Read: AI Leading The Evolution of Night Vision Camera Technology
The system will use anomaly detection algorithms to generate a consolidated view of likely threats ranging anywhere from a naturally emerging disease outbreak to that of a bioterrorism event. The first step would be to locate diversions from the relevant existing data, such as local air quality, viral presence in wastewater, information from home devices such as Alexa, hospital attendance figures, crowd sizes, and accident frequency.
How is it different from the current public health monitoring systems?
Existing public health monitoring systems mainly rely on manual reporting and are more reactive than proactive. This is a vision to try to change that into real-time data that can be coupled with AI for anomaly detection. Automating the analysis of such information will enable faster alerts when irregularities are noted and faster response times. Imagine a network permanently analyzing environmental, medical, and social data and sending alerts when anything looks out of whack: that’s the kind of proactive defense-making being envisioned.
Major hurdles faced in making this vision a reality
The biggest barrier that remains is the integration of data sources. The above system will require data across multiple disparate data sources, including hospital records, environmental sensors, and public health databases. Combining such diverse data into one understandable, safe format is extremely difficult. Another concern would be privacy issues. It is inevitable that medical data will cause privacy concerns.
So, it is paramount that we build a system that will get some insight while not ruining the privacy of the person. Lastly, releasing and building trust and collaboration is an important prerequisite to the success of this solution between various institutions, governments, hospitals, and research organizations.