A5007389371- Arctic and Antarctic ice dynamics
- Cryospheric studies and observations
- Underwater Vehicles and Communication Systems
- Oceanographic and Atmospheric Processes
- Climate change and permafrost
- Aerospace Engineering and Energy Systems
- Ocean Waves and Remote Sensing
- Modular Robots and Swarm Intelligence
- Neural Networks and Applications
- Neural dynamics and brain function
- Robotics and Sensor-Based Localization
- Robotic Path Planning Algorithms
- Hand Gesture Recognition Systems
- Advanced Electrical Measurement Techniques
- Biomedical and Engineering Education
- Underwater Acoustics Research
- Advanced Memory and Neural Computing
- Human Pose and Action Recognition
- Indoor and Outdoor Localization Technologies
- Coastal and Marine Dynamics
- Hearing Impairment and Communication
- Robotic Locomotion and Control
- Speech and dialogue systems
- Power Quality and Harmonics
- Magnetic Properties and Applications
University of Cape Town
2012-2024
Abstract. Two ensembles of buoys, deployed in the marginal ice zone (MIZ) north-eastern Weddell Sea region Southern Ocean, are analysed to characterise dynamics driving sea drift and deformation during winter-growth spring-retreat seasons 2019. The results show that although two buoy arrays were within same ice-covered ocean, their trajectory patterns vastly different. This indicates a varied response each season local winds currents. Analyses winter data showed Antarctic Circumpolar Current...
This research addresses the Simultaneous Localization and Mapping (SLAM) problem in context of an underground mining environment. would allow autonomous vehicle navigation this hazardous setting. environmental setting has few features no access to GPS or stationary towers, which are typically used for mobile robots. In addition, dust debris may hinder optical methods ranging. study investigates using randomly distributed RFID tags autonomously localize navigate such environments. is...
In this study learning reinforcement and noise rejection of a spatial pooler was examined, the first stage in Hierarchical Temporal Memory (HTM) network. is proposed model within field neuromorphic engineering. It describes top down approach to understanding how human brain performs higher reasoning has application as machine-learning algorithm. Final results displayed an increase permanence values associated with input pseudo-sensory signal system able accurately recognize up twenty percent...
This paper proposes a novel approach to estimating the sea state and hence predicting short-term, local wave behaviour from inertial measurements on slow-moving marine platform. A Kalman filtering strategy consisting of phase-locked loop filter based estimator is used generate height angle arrival estimates. method offers an improvement over existing Fast Fourier Transform methods as it does not require long time data series produce results, enables prediction short into future. The ideas...
This paper discusses a metering concept design based on equipment cheaper and smaller than laptop able to meet the requirements of power measurements that need relatively high frequency, simultaneous sampling. Its size cost make it useful when developing hardware for smart grid systems. Non-active is term used describe both distortion reactive power. In conventional theory, apparent power, S, non-active Q, are RMS voltages line currents. A novel general theory was proposed recently give...
Abstract. Two ensembles of buoys, deployed in the north-eastern Weddell Sea region Southern Ocean, are analysed to characterise dynamics driving sea ice drift and deformation during winter-growth spring-retreat seasons 2019. The results show that although two buoy arrays were within same ice-covered ocean, their trajectory patterns vastly different. This indicates a varied response each season local winds currents. Analyses winter data showed Antarctic Circumpolar Current modulated near...
Six ice-tethered instruments were deployed in the Antarctic Marginal Ice Zone during July 2022, coinciding with movement of a polar cyclone through this region. High frequency inertial and GPS location data retrieved from instruments. Inertial analysed two to investigate waves-in-ice dynamics ice floes on which deployed. In particular, large frequent impacts between detected time-series. This turn has implications for wave-ice interaction attenuation, wave measurement calibration remote sensing data.
This paper presents a preliminary investigation into the 3D reconstruction of pancake sea ice using multi-sensor approach with LiDAR (Livox Avia) and depth cameras (Intel Realsense D455). Pancake is type formed in dynamic region Antarctic Marginal Ice Zone (MIZ). Current models simplify floe geometry, neglecting effects geometry has on mobility ice. A series scans were taken SCALE Winter 2022 expedition where static was explored customised imaging rig. pipeline for processing data advantages...
Deaf and hard of hearing (DHH) people are estimated to exceed 5% the total population South Africa world. Additionally, majority with abilities not proficient in sign language. This creates a communication divide between DHH communities, disadvantage DHH. Although some provisions made cater needs form language education interpreters, such interpreters severely scarce often costly. In this research, we developed first vision based neural translation (NSLT) model for African Sign Language...
The remote navigation of small surface marine vehicles is an exciting engineering challenge. It has applications in the coordination fleets unmanned vessels that are used observation and data gathering mission our oceans. However, wave field region, where operating, required for a reliable resilient navigation. In this initial study, we present Kalman Filter method derived from lossless equation to estimate sea state generate local applications. A numerical case study was simulated,...