As billions of connected devices become part of everyday life, protecting them has become one of cybersecurity’s biggest challenges. From smart homes and hospitals to factories and critical infrastructure, these devices often rely on security systems that either follow fixed rules or depend on distant cloud servers to detect threats, approaches that can be too slow for attacks that unfold in real time.

For cybersecurity researcher Bright Gazie Akwaronwu, the solution is to make devices intelligent enough to defend themselves.

An Assistant Lecturer at Babcock University, where he teaches cybersecurity and digital electronics, Akwaronwu has focused his recent research on moving security closer to where attacks occur. Rather than waiting for remote systems to analyse suspicious activity, he believes connected devices should be capable of identifying and responding to threats independently.

That philosophy underpins one of his latest research projects, developed with Ajaegbu Chigozirim, Adediran Oluwaseyi Segun and Bamikole Olarewaju Aina. The team designed a wireless, context-aware intrusion detection system that operates directly on edge devices.

Instead of relying solely on predefined rules, the system analyses contextual information, including device behaviour, wireless communication patterns and operating conditions, to determine whether an activity is legitimate or malicious. Running on low-power technologies such as Wi-Fi, Bluetooth and ZigBee, it is designed to detect suspicious behaviour in real time and automatically respond by blocking malicious traffic or isolating compromised devices before attacks can spread.

According to the researchers, testing showed the system consistently detected intrusion attempts while keeping false alarms to a minimum, an important consideration for environments where unnecessary alerts can overwhelm administrators.

In a separate study, Akwaronwu collaborated with E. C. Ogu, Innocent U. Akwaronwu, Oluwabamise J. Adeniyi and Ayodeji G. Abiodun to examine how machine learning is being applied to protect modern cryptographic systems.

Rather than developing a new model, the team reviewed existing research to evaluate which artificial intelligence techniques perform most effectively against attacks targeting encryption systems. Their findings suggest that while established machine learning algorithms remain highly effective, newer hybrid approaches are beginning to deliver even stronger performance in detecting increasingly sophisticated threats.

Beyond cybersecurity, Akwaronwu has also explored the application of machine learning in healthcare. Earlier this year, he co-authored two systematic reviews examining how artificial intelligence can improve the prediction of tuberculosis drug resistance and lung cancer risk, reflecting the growing role of data-driven technologies across multiple disciplines.

His growing research profile has also seen him contribute to the wider academic community. In April, he served as a peer reviewer for an international conference on information and communication technology, assessing research submitted by other scholars before publication.

Akwaronwu completed his Master’s degree in Computer Science, specialising in Cybersecurity, with distinction. Before joining academia, he spent six years managing the ICT infrastructure of a Nigerian secondary school, an experience he says continues to shape his practical approach to designing secure digital systems. He also holds a professional certification in Telecommunications and Network Security from the Computer Professionals Registration Council of Nigeria.

His current research continues to build on the same theme. Among the projects he is pursuing are the use of isolation-based algorithms to detect advanced persistent threats within network traffic, methods for verifying hardware integrity after semiconductor manufacturing, and studies examining trust and security across modern chipset supply chains.

As the number of connected devices continues to grow worldwide, Akwaronwu believes cybersecurity must evolve beyond reactive defence.

His work points towards a future in which connected devices no longer wait for remote servers or human intervention before responding to threats, but instead analyse, decide and act autonomously, bringing intelligence to the very edge of the network where security matters most.