The development of the Internet of Things (IoT) has shown significant contributions to many
application areas, such as smart cities, smart homes, and smart farming, including aquarium control systems.
Important things in an aquarium system are the level of ammonia in the water and the temperature of the
water. Other research proposes several systems to make the aquarium control system robust for the aquarium
monitoring and control system. However, those systems have weaknesses; namely, the user must actively
access information to the server. This paper proposes a robust aquarium control system using the decision
tree regression (DTR) algorithm. The development of this system was to overcome the problem of aquarium
control by remote users. An accurate and real-time system is needed to monitor the aquarium so that it does
not reach dangerous and critical points, such as in the case of an increase in water temperature. We did tests
by developing an aquarium system connected to a server and an application that acts as a controller. Our
measurements check the delay of sending data from the sensor to the server, process delay, actuator delay,
user delay, and delay in reaching the aquarium’s critical point. The measurement of the system’s robustness is
by calculating the probability of the information arrival to the user in the period of the critical point compared
to the time needed to reach the critical point. Furthermore, we also made an analytical model based on the
probability density function of the delay covered in this system. Analytically and experimentally, we show
that the system can meet the needs of aquarium monitoring and control in an IoT-based environment.