“The goal is to develop a robust solution to tackle the growing threat from cybercrime, especially in the health sector where it is crucial to ensure effective monitoring and alerts on irregularities in patient care,” says Per-Arne Andersen, associate professor at the University of Agder (UiA).
The duration of the project is four years and it will run from 2024 to 2028. It will be led by UiA's Associate Professor Per-Arne Andersen and will be carried out in collaboration with researchers from the Hospital of Southern Norway and Newcastle University.
“We will develop systems for cyber monitoring and alerts using the Tsetlin machine method of using artificial intelligence,” Andersen says.
The researchers will design a system that uses minimal power to predict attacks on so-called Internet- of-Things-devices (IoT devices).
IoT devices can be devices that are connected to the internet, such as lamps, household appliances, industrial machinery and medical equipment. The devices have limited processing power, memory and battery capacity.
Logical and energy efficient
The Tsetlin machine is a type of machine learning algorithm or artificial intelligence code. It was developed by UiA professor Ole-Christoffer Granmo.
The solution is known for handling complex problems efficiently and with low energy consumption. This makes the method ideal for use in devices with limited resources such as IoT devices.
In addition, the Tsetlin machine performs calculations based on logic. This means that it is easy for researchers to understand how the artificial intelligence arrives at its solutions.
Monitoring data devices
“We will explore techniques for using the Tsetlin machine to actively analyse IoT systems and detect irregularities, either in the form of hacking or other errors that make continued use risky,” says Andersen.
According to Andersen, this is just the beginning. The long-term goal is an even larger research project financed by EU funds.
Andersen is expected to envisage the entire process from testing drafts and prototype before a final version of the Tsetlin machine security system can be put to use.
“Currently, there are no effective systems that monitor data devices and detect irregularities. Now we’re going to develop a system where all IoT devices, for example at a hospital, get an active monitoring mechanism that sends alerts if something is wrong. The system acts as some sort of a watchdog,” says Andersen.
He hopes that this will significantly improve security, especially in cases where malicious actors manage to break passive security measures such as encryption and authentication.
Collaboration with hospital and Newcastle
The Research Council has allocated NOK 15 million to UiA's research project called "SecureloTM: Ultra-low-energy IoT Intrusion Detection Systems using Logic-Based Tsetlin-Machines".
At UiA, the project will employ three PhD research fellows and one postdoctoral researcher.
Among the other researchers on the project are Andersen’s former supervisors for his doctoral work, professors Ole-Christoffer Granmo and Morten Goodwin.
Other important partners are, as previously mentioned, Newcastle University and researchers Geir Thore Berge and Tor Oddbjørn Tveit from the Hospital of Southern Norway. The latter have previously carried out several research projects in collaboration with UiA.
“The goal is to develop a completely new type of security mechanism that will be able to operate with high precision, while power consumption is kept to a minimum,” Andersen says.