Hayati, H, Walker, P, Mahdavi, F, Stephenson, R, Brown, T & Eager, D 2018, 'A Comparative Study of Rapid Quadrupedal Sprinting and Turning Dynamics on Different Terrains and Conditions: Racing Greyhounds Galloping Dynamics', Volume 4A: Dynamics, Vibration, and Control, ASME 2018 International Mechanical Engineering Congress and Exposition, ASME, Pittsburgh, Pennsylvania, USA, pp. 1-7.View/Download from: UTS OPUS or Publisher's site
Identifying optimum athletic race track surfacing for greyhounds to reduce risk of injuries is a challenging practice as there are several single and coupled variables that should be considered as risk factors. To study the impact of bend and straight sections, surface type and camber, on biomechanics of galloping quadrupeds, an inertial measurement unit (IMU).
has been used to measure the associated galloping accelerations. The IMU was sewn into a pocket located on the back of the greyhounds racing jacket positioned between the two forelegs. Simultaneous kinematics were performed using high frame rate (HFR) videos for calibrating IMU data. The results showed that there were lower G-forces on galloping on grass than wet sand which is consistent with the mechanical behavior of grass (grass is softer than wet sand). Moreover, galloping around the bend had higher G-forces than galloping along the straight section suggesting an excessive force is applied on the greyhound's limbs due to centrifugal force. A cambered bend assisted the greyhounds in having a smoother gait and lower G-forces when compared to a flat bend. The results reported in this paper will not only be beneficial for the welfare of racing greyhounds, but will also contribute in the simulation of legged locomotion for bio-inspired engineering and robotics.
Stephenson, RM, Chai, R & Eager, D 2018, 'Isometric Finger Pose Recognition with Sparse Channel SpatioTemporal EMG Imaging.', 2018 40th Annual International Conference of the IEEE Engineering in Medicine and Biology Society. IEEE Engineering in Medicine and Biology Society Proceedings, Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBS) Conference, IEEE, USA, pp. 5232-5235.View/Download from: UTS OPUS or Publisher's site
High fidelity myoelectric control of prostheses and orthoses isparamount to restoring lost function to amputees and neuro-muscular disease sufferers. In this study we prove that patio-temporal imaging can be used to allow convolutional neural networks to classify sparse channel EMG samples from a consumer-grade device with over 94 % accuracy. 10,572 images are generated from 960 samples of simple and complex isometric finger poses recorded from 4 fully intact subjects. Real-time classification of 12 poses is achieved with a 250ms continuous overlapping window.
Hayati, H, Eager, D, Stephenson, R, Brown, T & Arnott, E 2017, 'The impact of track related parameters on catastrophic injury rate of racing greyhounds', 9th Australasian Congress on Applied Mechanics, ACAM 2017, Australasian Congress on Applied Mechanics, Engineers Australia, Sydney, Australia.View/Download from: UTS OPUS
© 2017 National Committee on Applied Mechanics. All Rights Reserved. Greyhounds can travel twice as fast as human athletes, attaining constant average running speeds of ~65 km/h vs ~29 km/h. Their locomotion is also different from human sprinters, and more similar to cyclists. Unlike human sprinters where the muscles powering the locomotion are also supporting the weight, locomotion of greyhound are powered by torque about the hip. Agile, high-speed quadrupeds, such as the greyhound, experience extreme ground-limb contact forces while negotiating turns; leading to an increased susceptibility to injuries. Added to this, rapid, high velocity changes in direction and extreme turning angles magnify the lateral acceleration forces experienced on the limbs and torso. In this paper, the rate of severe musculoskeletal injuries of racing greyhounds at 34 tracks in New South Wales, Australia, were obtained for the year of 2016. The correlation of parameters, namely bend radius, bend camber, bend length and back straight length and the catastrophic injury rate are statistically analyzed . Track injury locations were obtained from race video footage No correlation was seen between catastrophic injury rate and bend radius, bend camber, bend length and back straight length. Analyses revealed the highest injury rate based on location to be at the first turn. Footage lends support to this being caused by the immediate clustering of the greyhounds towards the inner 'lure' rail.' The results of this study support previous findings that greyhounds racing in an anti-clockwise direction most commonly suffer musculoskeletal injuries to their right hind limbs which is consistent with knowledge of the forces that occur on the leading limbs of these dogs as they maintain their speed around bends.
Stephenson, RM, Naik, GR & Chai, R 2017, 'A System for Accelerometer-Based Gesture Classification Using Artificial Neural Networks', Proceedings of the 39th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC'17), Annual International Conference of the IEEE Engineering in Medicine and Biology Society, IEEE, Jeju island, Korea, pp. 4187-4190.View/Download from: UTS OPUS or Publisher's site
A great many people suffer from neurological movement disorders that render typical hardware interface devices ineffective. A need exists for a universal interface device that can be trained to accept a wide range of inputs across varying types and severities of movement disorders. In this regard, this paper details the design, testing and optimization of an accelerometer-based gesture identification system. A Bluetooth-enabled IMU mounted on the wrist provides hand motion trajectory information to a local terminal. Several techniques are applied to decrease the intra-class variance and reduce classifier complexity including filtering, segmentation and temporal scaling. Datasets consisted of 520 training samples, 260 validation samples and a further 520 testing samples. A multi-layer feed forward artificial neural network (ML-FFNN) was used to classify the input space into 26 different classes. Initial system accuracy, using arbitrary hyperparameters was 77.69% with final optimized accuracy at 99.42%.
A summary brief on current track curation ideology and procedures as well as the proposal of a new LIDAR-based data driven approach. Prepared for and presented to Greyhound Racing Victoria.