SUMMARY: Tethered mobile robots are ideal for electrically noisy environments and for time-consuming tasks that require robust data communication and uninterrupted power delivery. However, tethers may become entangled in cluttered environments, leading to immobilisation and consequent mission failure. This work addresses real-time monitoring of tethers to detect tether entanglement, perform disentanglement through tether following and localise within line of sight. Experimental hardware is proposed to implement the tether monitoring techniques. Experiments are performed for single and dual mobile robots to search a target environment and entanglement detection is shown to be successful using quantitative metrics such as mean localization error.

High-speed photography and image processing were used to study the aerodynamics in the cylinder–doffer transfer region. Smoke tracers were utilised to delineate aerodynamic patterns under the cylinder–doffer set point and the images were analysed as an indication of the pressure and velocity changes of the air flow in the region. High-speed photographic observations of fibre motion in the region were also studied in order to determine the effects of the air flow on fibre transfer. The results showed that due to local air pressure differences, air turbulence is generated under the cylinder–doffer set point, which may affect fibre configuration during the transfer. From the results an explanation is proposed for the occurrence of turbulent flow under the cylinder–doffer set point, and ideas previously reported by other researchers about the fibre transfer mechanism are further developed.

Falls are a major health issue in societies globally and the second leading cause of unintentional death worldwide. To address this issue, many studies aim to remotely monitor gait to prevent falls. However, these activity data collected in studies must be labelled with the appropriate environmental context through Human Activity Recognition (HAR). Multimodal HAR datasets often achieve high accuracies at the cost of cumbersome sensor systems, creating a need for these datasets to be analysed to identify the sensor types and locations that enable high-accuracy HAR. This paper analyses four datasets, USC-HAD, HuGaDB, Camargo et al.’s dataset, and CSL-SHARE, to find optimal models, methods, and sensors across multiple datasets. Regarding window size, optimal windows are found to be dependent on the sensor modality of a dataset but mostly occur in the 2–5 s range. Support Vector Machines (SVMs) and Artificial Neural Networks (ANNs) are found to be the highest-performing models overall. ANNs are further used to create models trained on the features from individual sensors of each dataset. From this analysis, Inertial Measurement Units (IMUs) and three-axis goniometers are shown to be individually capable of high classification accuracy, with Electromyography (EMG) sensors exhibiting inconsistent and reduced accuracies. Finally, it is shown that the thigh is the optimal location for IMU sensors, with accuracy decreasing as IMUs are placed further down away from the thigh.