Brain-computer interface for hands-free teleoperation of construction robots

Recently, the use of collaborative robots has started to emerge at construction sites. Such incorporation into these human-dominated environments can raise safety concerns as most robots are not fully automated and require some sort of control mechanism. Conventional control systems may fall short under specific situations in which workers require robotic assistance but cannot use their hands to control the robot. Brain-computer interface (BCI) can offer such hands-free controllability, a non-muscular communicative channel that can establish an interpretive pathway between humans and robots. This paper proposes a BCI-based system to remotely control a robot by continuously capturing workers' brainwaves acquired from a wearable electroencephalogram (EEG) device and interpreting them into robotic commands with 90% accuracy. The findings revealed the proposed system holds promise for enhancing robot control in hazardous operations where the ability of the worker to physically direct the robot is limited, such as underwater and space construction. •A Motor imagery (MI)-driven brainwave-computer interface (BCI) system was proposed to enable hands-free control of robots.•The proposed system generates robotic commands by translating brainwaves related to imagination.•The accuracy of the predictive models, produced by an ensemble classifier, was almost 80% with an over-90% confidence level.•The proposed system can be used in hazardous environments with limited worker's movability.