STS Defence Ltd.

United Kingdom

STS Defence Ltd.

United Kingdom
SEARCH FILTERS
Time filter
Source Type

Yusuf S.A.,STS Defence Ltd. | Brown D.J.,University of Portsmouth | Mackinnon A.,STS Defence Ltd.
Robotics and Autonomous Systems | Year: 2015

Audio event detection (AED) and recognition is a signal processing and analysis domain used in a wide range of applications including surveillance, home automation and behavioral assessment. The field presents numerous challenges to the current state-of-the-art due to its highly nonlinear nature. High false alarm rates (FARs) in such applications particularly limit the capabilities of vision-based perimeter monitoring systems by inducing high operator dependence. On the other hand, conventional fence-based vibration detectors and pressure-driven "taut wires" offer high sensitivity at the cost of a high FAR due to debris, animals and weather. This work reports an audio event identification methodology implemented as a test-bed system for a surveillance application to reduce FAR, maximize blind-spot coverage and improve audio event classification accuracy. The first phase utilizes a nonlinear autoregressive classifier to locate and classify discrete audio events via an exogenous sound direction variable to improve classifier confidence. The second phase implements a time-series-based system to recognize various audio activity groups from nominal everyday sound events such as traffic and muffled speech. The discretely labeled data is thus trained with HMM and Conditional Random Field classifiers and reports a substantial improvement in classification accuracies of indoor human activities. © 2015 Elsevier B.V. All rights reserved.


Syed Y.A.,University of Portsmouth | Brown D.J.,University of Portsmouth | Garrity D.,STS Defence Ltd | Mackinnon A.,STS Defence Ltd
2015 5th National Symposium on Information Technology: Towards New Smart World, NSITNSW 2015 | Year: 2015

Pedestrian navigation via dead reckoning (PDR) is considered a promising domain for search and rescue personnel tracking, particularly for fire-fighters. The technique is considered particularly useful when other conventional means such as the GPS and RF-based location estimation are not present or not accurate. However, PDR approaches in real-world operating environments fail due to a wide range of factors ranging from the personnel's natural behavior to diversity of activities a first-responder may perform during a rescue mission. This technique presents a PDR activity classification technique utilizing shoe-mounted microelectromechanical sensors for efficient step and attitude analysis via a 2D Kalman filter. The methodology then utilizes HMMs for various activity types such as walking, side-stepping, crawling, etc. Tests performed on the proposed technique showed the step identification technique to perform well with an overall accuracy of 90.75% in step-counting where a simple Naïve Bayes classifier was used. The HMM-based activity classifier presented 86% and 85% accuracy in correctly identifying upstairs and downstairs walking activity. © 2015 IEEE.


Kazemian H.,London Metropolitan University | Yusuf S.A.,STS Defence Ltd | White K.,London Metropolitan University | Grimaldi C.M.,London Metropolitan University
Expert Systems with Applications | Year: 2016

This paper is concerned with applications of a dual Neural Network (NN) and Support Vector Machine (SVM) to prediction and analysis of beta barrel trans membrane proteins. The prediction and analysis of beta barrel proteins usually offer a host of challenges to the research community, because of their low presence in genomes. Current beta barrel prediction methodologies present intermittent misclassifications resulting in mismatch in the number of membrane spanning regions within amino-acid sequences. To address the problem, this research embarks upon a NN technique and its comparison with hybrid-two-level NN-SVM methodology to classify inter-class and intra-class transitions to predict the number and range of beta membrane spanning regions. The methodology utilizes a sliding-window-based feature extraction to train two different class transitions entitled symmetric and asymmetric models. In symmetric modelling, the NN and SVM frameworks train for sliding window over the same intra-class areas such as inner-to-inner, membrane(beta)-to-membrane and outer-to-outer. In contrast, the asymmetric transition trains a NN-SVM classifier for inter-class transition such as outer-to-membrane (beta) and membrane (beta)-to-inner, inner-to-membrane and membrane-to-outer. For the NN and NN-SVM to generate robust outcomes, the prediction methodologies are analysed by jack-knife tests and single protein tests. The computer simulation results demonstrate a significant impact and a superior performance of NN-SVM tests with a 5 residue overlap for signal protein over NN with and without redundant proteins for prediction of trans membrane beta barrel spanning regions. © 2016 Elsevier Ltd


Yusuf S.A.,STS Defence Ltd | Brown D.J.,University of Portsmouth | Mackinnon A.,STS Defence Ltd | Papanicolaou R.,STS Defence Ltd
Proceedings of the International Joint Conference on Neural Networks | Year: 2013

Intelligent assessment of information gathered from industrial-grade data loggers for preemptive maintenance is one of the foremost areas of research in conditional monitoring. Due to the general operating environment, there exists a non-linear relationship between the input and output data gathered from these sensors. Moreover, the transmission of data from such dynamic environments is generally marred by a large SNR with substantial level of 'false-noise' belonging to the normal movement pattern of the mechanical parts. Within this context, the goal of this paper is to explore, evaluate and develop an optimal, dynamic neural network to improve the fault prediction accuracy of condition monitoring systems. The training data for this research was obtained from a vibration and a thermal sensor connected mounted over a polyphase induction motor. The objective was to identify any anomalies in the motor's fan-based cooling system. Moreover, the model presented a comparative analysis of a dynamic neural network (DNN) model against a non-linear autoregressive neural system (NARX) with exogenous input. The validation outcome presented a close regressive relationship of 0.9734 between observed and targeted outcomes over a 7-second delay with a NARX model giving a 4.56% and 5.23% classification accuracy. The best model system was evaluated against unseen anomaly data and demonstrated high prediction accuracy. © 2013 IEEE.


Yusuf S.A.,STS Defence Ltd | Brown D.J.,University of Portsmouth | Mackinnon A.,STS Defence Ltd | Papanicolaou R.,STS Defence Ltd
Proceedings of the International Joint Conference on Neural Networks | Year: 2013

Background subtraction is a well-known technique in computer vision to extract foreground objects from background reference frames. In real-time video processing applications such as surveillance, behavioral profiling and intelligent transport systems, the domain presents a number of challenges. Video frames used to train such models contain a range of dynamic background activities such as waving trees, moving cloud cover or abrupt intensity variations that make the foreground detection a challenging task. Dynamic neural networks are known for their capability to predict time-series-based nonlinear models via previous feature data. The proposed scenario models each pixel's intensity/color-alternating behavior based on its previous activity patterns. Any significant or unusual variation in the underlying intensity or color value therefore is modeled as a foreground activity. Based on this concept, this paper presents a non-linear autoregressive neural (BG-NARX) classifier with the pixels' chromatic values as the exogenous vectors to improve background detection accuracy. The proposed model was evaluated against three benchmarking video datasets and reported promising detection accuracies ranging from 67-94% for pedestrians and vehicles against highly variable backgrounds with low false positives and negatives. © 2013 IEEE.


Yusuf S.A.,STS Defence Ltd | Brown D.J.,University of Portsmouth | Mackinnon A.,STS Defence Ltd
Robotics and Autonomous Systems | Year: 2015

Audio event detection (AED) and recognition is a signal processing and analysis domain used in a wide range of applications including surveillance, home automation and behavioral assessment. The field presents numerous challenges to the current state-of-the-art due to its highly nonlinear nature. High false alarm rates (FARs) in such applications particularly limit the capabilities of vision-based perimeter monitoring systems by inducing high operator dependence. On the other hand, conventional fence-based vibration detectors and pressure-driven "taut wires" offer high sensitivity at the cost of a high FAR due to debris, animals and weather.This work reports an audio event identification methodology implemented as a test-bed system for a surveillance application to reduce FAR, maximize blind-spot coverage and improve audio event classification accuracy. The first phase utilizes a nonlinear autoregressive classifier to locate and classify discrete audio events via an exogenous sound direction variable to improve classifier confidence. The second phase implements a time-series-based system to recognize various audio activity groups from nominal everyday sound events such as traffic and muffled speech. The discretely labeled data is thus trained with HMM and Conditional Random Field classifiers and reports a substantial improvement in classification accuracies of indoor human activities. © 2015 Elsevier B.V.


Grant
Agency: GTR | Branch: Innovate UK | Program: | Phase: Feasibility Study | Award Amount: 111.50K | Year: 2014

Project CLAIMS (Coolant Leak Artificially Intelligent Monitoring System) will provide a technology demonstrator, of advanced automated condition monitoring system for detection and classification of leaks from the primary circuits of designs of light water reactors. Nuclear plants are valuable, high capital cost assets with long operating lifetimes delivering reliable base electricity load to the grid. It is recognised that in order to maintain optimal safety and economic viability, application of advanced Surveillance, Diagnostic and Prognostic technologies will be required, particularly as plant lifetimes extend to 80 years and beyond.


Grant
Agency: GTR | Branch: Innovate UK | Program: | Phase: Smart - Development of Prototype | Award Amount: 148.90K | Year: 2014

Project FiRST (Fire-fighter Rescue and Support Technology) will produce a pre-production prototype of a novel Personal Protective Equipment (PPE) concept, that provides a stepchange in fire-fighter situational awareness when tackling fires in a range of environments, but particularly those challenges represented by fighting fires indoors. Fire-fighters tackling indoor blazes face exceptional hazards; from direct heat/hot gases produced from burning materials, structural instability, through to severely impaired situational awareness due to poor visibility, noise, and the difficulties Incident Commanders have in knowing the location and status of their teams inside larger structures. These particular circumstances have combined in past incidents, together with sudden and catastrophic rises in temperatures, to directly lead to fire-fighter fatalities. STS Defence is employing its extensive expertise to develop a prototype system to support front line Fire & Rescue Services. It is non-invasive, integrated onto existing breathing apparatus sets, and provides increased awareness of the fire-fighter to his/her surrounds. A robust radio-frequency communications link back to the Incident Commander provides the data and therefore the means of identifying the status of the teams, and their location in GPSdenied indoor locations.


Grant
Agency: GTR | Branch: Innovate UK | Program: | Phase: Innovation Voucher | Award Amount: 5.00K | Year: 2013

STS Defence are a Gosport-based SME, providing contract manufacturing & through-life system engineering services; engaging with our customers through experience, expertise and a commitment to quality.STS Defence had a turnover of £12.1m in 2012.We focus on working in partnership with customers developing initial concepts into reality, leveraging over 40 years’ experience of working on Ministry of Defence projects to develop an impressive portfolio of through-life capabilities that are essential to our clients.Our offer is based on flexibility, agility and quality, with a breadth of design-led manufacturing services for electronic and electro-mechanical assembly from prototyping & procurement to fabrication, finishing and test while our through-life systems engineering services range from design & integration to installation & logistic support.


Grant
Agency: GTR | Branch: Innovate UK | Program: | Phase: Collaborative Research & Development | Award Amount: 1.03M | Year: 2015

Project IConIC (Intelligent Condition monitoring with Integrated Communications) will develop a revolutionary system that provides accurate prediction for vessel performance and operation. It represents a step-change in the marine industry, through linking autonomous measurement systems to existing maritime satellite communications. The performance of marine engines used for propulsion and power generation has a significant impact on efficient vessel operation. Inefficient/failed engines on large vessels can cost their owners £millions. In response to this, STS Defence, a Gosport based SME, has gathered together a strong consortium of companies and academic partners to develop an automated machine to machine and ship-to-shore data exchange capability, increasingly described as the internet of things.

Loading STS Defence Ltd. collaborators
Loading STS Defence Ltd. collaborators