A member of Ashesi’s faculty since 2004, Associate Professor Nathan Amanquah has taught computer science and engineering courses focusing on smart and intelligent systems. He has now been appointed Dean of Engineering and Computer Science after serving as acting Dean since 2018. Combining his passion for teaching and research, Prof. Amanquah recently published a textbook titled “Learning to Design Embedded Systems.” The book covers three strategies for designing smart systems: manipulating individual bits directly, using an operating system, and integrating low-level assembly code for better speed. In the essay and interview below, he shares why smart systems will improve the world.
Article written by Associate Professor Nathan Amanquah
Digital electronics are now everywhere, and we enjoy digital versions of anything analogue. We use these devices daily, from microwaves to blood pressure monitors, anti-lock braking systems (ABS) in cars, and even the MARS rover. These systems are getting smarter, smaller, and used in more areas. Intelligent technologies can monitor the physical environment and act without human help. For example, smart buildings can save energy by turning off unused services.
I love motivating my students to build and implement these technologies. In many of my classes at Ashesi, we cover the basics of designing anything digitally intelligent. This includes devices using "combinational" logic and those with stored memory, like vending machine controllers or elevator controllers. In my Embedded Systems course, we build systems using low-power computing platforms called microcontrollers.
Microcontrollers are the foundation of smart systems. Students learn to program these at a low level to make them responsive, reliable, and power-efficient. On the other hand, I teach students about the "Internet of Things" (IoT), where we use smart technology to monitor various phenomena. The data collected can give insights that help systems act independently as needed.
My research also focuses on improving smart devices. Often, smart objects rely on remote internet servers for AI-related tasks, which can cause privacy, security, and delay issues. My research aims to make smart devices better by bringing the AI tasks onto the device itself, solving these problems and improving responsiveness. For example, we have used embedded machine learning to estimate water flow rates in pipes without cutting them open. Similarly, we are studying driving patterns to build a smart system that offers drivers better insurance rates if they have fewer incidents. Smart embedded systems will become even more advanced, and those who study and research them can use them to improve the world.
