Table of contents

With great honour, we would like to welcome all delegates to our 2019 International Conference on Man Machine Systems (ICoMMS). It is a great pleasure and privilage for the Mechatronic Engineering Programme, under the School of Mechatronic Engineering, Universiti Malaysia Perlis (UniMAP) to organize this conference.

ICoMMS conference covers a wide range of current research areas all of which are important elements in the related engineering fields. This also reflects the diverse research interests conducted under the School. This conference gathers experts from different countries to discuss and present the state of the art and new research findings and perspectives. We are sure all participants will use this conference as a platform to present their work and address current problems in this challenging but promising field and subsequently create new avenues for the commercialization of the research outcomes.

We would like to express our sincere gratitude to all reviewers, both local and international, for their time and effort in reviewing the papers. Their feedbacks to the authors have been excellent and useful to ensure the high quality of papers for the conference. Our sincere thank also goes to all sponsors for their generosity. Last but not least, we very much appreciate and thank UniMAP and all members of the organizing committee for their hard work, dedication, cooperation and suggestions in ensuring the success of this conference.

Thank you very much and hope to see you again in our next conference.

Prof. Dr. Abdul Hamid Adom

Dr. Norasmadi Abdul Rahim

Dr. Lam Chee Kiang

Dr. Eng Swee Kheng

Dr. Kamarulzaman Kamaruddin

Editors

2019 International Conference on Man Machine Systems (ICoMMS)

All papers published in this volume of IOP Conference Series: Materials Science and Engineering have been peer reviewed through processes administered by the proceedings Editors. Reviews were conducted by expert referees to the professional and scientific standards expected of a proceedings journal published by IOP Publishing.

Moisture content (MC) is that the weight of water contained in paddy and other grans expressed in percent. Moisture content is typically mentioned the wet basis which means the whole weight of the grain including the water (MCwb). Accurate MC testing is very important in managing and selling paddy and rice because depending on the aim rice has completely different ideal MC. Inaccurate MC measurements lead to lower head rice once milled at wrong moisture content, additional drying cost and loss of quality if paddy is dried too much, spoilage if the grain is just too wet in storage, additional drying cost and harvesting loss if paddy is harvested wetter than necessary and Weight loss (loss in profit) if grain is sold too dry. This paper intends to review the available techniques of moisture content measurements, its principles and sources of inaccuracies. Relevant properties of cereal grain also are studied with their impact on moisture content, temperature, density and humidity. There are several ways of grain moisture detection such as drying technique, resistance technique, capacitance technique, microwave technique, neutron technique, V/F Conversion, Ftir spectroscopic technique and Microcontroller technique. Current MC of grains are based- on sampling and individual single point measurement, so far, there is no available overall measurement of MC in silo, but by using tomography method can measurement MC in whole silo and investigate the efficiency and accuracy of localization of moisture distributions sensing in agricultural silo by using Gaussian algorithm.

A brief review on the dielectric properties of grain and their use in sensing moisture content is presented. Dielectric constant and loss factor were defined generally, and their correlation with grain permittivity from historical studies is presented. References are cited on dielectric measurement techniques using various frequency ranges from radio frequency through microwave frequencies. The described techniques including from the past decades until current studies.

Service robot is becoming more popular in recent study. It has opened up various opportunity to solve problems as well as the challenges need to be studied. While there are many successfully developed service robot out there, the general design principle with respect to the application need a description with solid justification. This review paper seeks and compare the design principle, specification, and hardware selection of several successfully developed service robot such as Sacarino and Blackbot. Based on the study, the most important design principles include simple mobile platform, modular structure, and aesthetic design.

In this paper, we present 3D point cloud data segmentation from a Terrestrial Laser Scanner (TLS) using RANSAC and localization approaches. Our work is using the real world data acquired from outdoor and indoor scene of Tawau Bell Tower (also known as the Belfry), one of the heritage building located in Sabah. We adapt the methods to segment ground and nonground data for data reduction. Based on the non-ground information and geometrical constraint, the data managed to show the building structure. The aim is to use and handle a smaller set of data rather than using the entire point cloud data in processing the region of interest. This will also help to reduce processing time as point cloud data from a TLS is usually large in size. The ground plane was fitted with RANSAC algorithm restricted by distance to describe the geometrical of the plane. As a result, the point cloud data was reduced to the environment around Bell Tower and able to highlight the region of interest. Building segmentation is then retrieved from the reference centre location to form a bounding box and data were organized using kd-tree to denote the Bell Tower structure. The framework is tested for outdoor and indoor scan of the Bell Tower and able to segment the data appropriately.

The aim of this project is to design and develop a face recognition based security system using machine vision. In this paper, a face recognition system using Principal Component Analysis (PCA) with Euclidean distance classifier is proposed. PCA is selected as it is less sensitive to noise and interference, besides reducing the number of variables in face recognition which reduce computational time and increase accuracy. A number of experiments were done to evaluate the performance of the face recognition system. A training database of 20 students is created and used in this project. Overall, the face recognition system on static face images has 86% success rate and 14% error rate which is mainly due to the variation of pose orientation. It is also found that the threshold value is an important factor for the performance of face recognition. The distance of the person to the camera play a role as the further the person is away from the camera, the more blur the face image captured. As for the anti-spoofing approach, the proposed detection of eye blinking is successfully tested with training eyes open and eyes close images.

Although many people with disabilities have a strong desire to play a musical instrument, there are a few support devices that allow them to play it. We propose four considerations that have been identified for the development of such a support device and developed "F-ready," allowing people with disabilities in their upper limbs to play the guitar. We demonstrate the results of a trial in which a male with a cervical spinal cord injury practiced playing the guitar with F-ready for five weeks. He could make sounds with the guitar using F-ready, even from his first day of practice. Then, he could simultaneously engage in three movements: moving his left hand to the appropriate position on the strings, pressing the F-ready switch, and strumming the strings with his right hand. After only five weeks, he performed a musical piece at a research meeting on the harmonica and guitar using F-ready. In addition, he faced several tasks: deciding on an optimum posture, how to deal with arm pain, and learning the song's chord progression. The experts solved these tasks together. The experiment demonstrated that F-ready supports players' ability to play the guitar at will and enthusiastically.

A fixed-wing UAV that is capable of vertical takeoff/landing is a hybrid aerial vehicle that could take off as helicopter and then transition into conventional airplane for forward flight. There are several powertrain configurations for the VTOL aircraft, one of them is called Separate Lift Thrust (SLT) configuration where the forward flight uses different powertrain compared to the hover phase. Without complex mechanism to store the inactive hover powertrain, the hover powertrain components added a significant amount of aerodynamic drag during forward flight. This paper presents the assessment of the drag caused by the inactive propellers during the forward cruising flight phase. For this research, 26 propeller samples with diameter from 5 inch to 9 inch were used with various configuration and materials. They were then tested in wind tunnel facility and the resultant drag was measured. The results from the wind tunnel test shows that for lowest drag penalty, smaller propeller diameter with low pitch blade provide the lowest drag. Interestingly, from the results, it shows that there is a flight speed that can provide an optimum drag from the chosen propeller. A selection of optimum motor and battery can be made in the future based on result presented in this paper to further improve the performance of the UAV.

Cloud computing is a type of computing that base on shared computer resources for delivering information technology services. Distributed Computational system has been considered as best systematic model for sharing resources. Minimization of response time is the requirement in the distributed computational system. Previous Algorithm belongs to the optimization in which every new Virtual Machine reaching as a request to the cloud computing manager. This creates disturbance in the processing during life cycle of others Virtual Machines. For overcome this problem, there is a proposed an algorithm for minimization response time in distributed computational system by using the multi agents cloud structure. The overall system is architecture as a distributed over multi layered model. This leads to a situation that demands high care in allocating a Virtual Machine (VM) to a server machine in order to improve allocation time and efficiency of resources. Focus of this work remains to decrease time of allocation to optimize response time using a Virtual Machine (VM) placement and allocation algorithm that accounts Virtual. Machines (VM), their types and available resources among the data center server machines. Presented algorithm is implemented by using cloud simulator tool, through this tool, the calculated results of presented algorithm have been quite better with respect to traditional algorithm, and the difference presented with graphs for the both algorithms. Results support proposed methodology.

Range of operation is one of the main limitation in the operation of Unmanned Aerial Vehicle (UAV). The ability to exploit currently deployed data-based cellular network for UAV application enables better utilization of UAV for various applications. Understanding the requirements of UAV application allows the identification of challenges to obtain coverage offered by terrestrial-based cellular networks. This paper outlines the challenges and identifies the relevance of cellular based UAV application in the context of Malaysia as well as its potential application. The main challenges highlighted include the design of cellular network which is optimized for terrestrial users. Additionally, in the context of Malaysia, the tropical weather imposes a further challenge to the reception of reliable radio signals. It has been identified that provided that these challenges can be overcome, the cellular based UAV can benefit the agriculture, security and utility sector in Malaysia.

Muscle fatigue is described by the decline in muscle maximum force during contraction. The fatigue occurs in the nervous or muscle fibre cells. The nerves produce a high-frequency signal to gain the maximum contraction, but it cannot sustain the high frequency signal for a long time, and that leads to a decline in muscle force. The surface Electromyography (EMG) is the dominant method to detect muscle fatigue because the EMG signals give more information about the muscle's activities. This review discussed the EMG signal processing and the methods of detection muscles fatigue with three domains (time domain, frequency domain, and time-frequency domain) based on EMG signals that are collected from the muscles during dynamic and static movements.

CAN-bus technology is adapted in various vehicles to reduce the wiring harness and controls the vehicle exploitation physical system. In this paper, Control Area Network (CAN)- bus protocol is enforced employing a Single Board Computer (SBC) and access with each other by set a sender and receiver to get the data transmission. A special CAN-bus application layer protocol is intended for the high reliable and high period management network of the vehicle system. However not all microcontroller that used for Computer module in vehicle network had CAN-bus communication port. Thus the CAN-bus Converter is used as tool for converting data from a computer into an understandable language. The CAN-bus converter can also be used in various ways to obtain the final results of the data sent to it. Although different methods may be used, the most appropriate method cannot be identified. This project is therefore designed to find and evaluate which method is suitable for obtaining accurate data results and for avoiding abnormal data and data transfer speed.

This paper focus on analysis on Long Range (LoRa) handover session was conducted in urban, suburban, and rural area in Perlis state, Malaysia. The experiment will consists of observation and analyzed on three area using LoRa communication modules. These three areas were collected on its Received Signal Strength Indicator (RSSI) to know the signal strength over the distance from base station and the differences physical object appear in Line of Sight (LoS) during the transmitting and receiving its signal from LoRa Transmitter to LoRa Receiver. The result at three area of different location have shown various pattern in term of RSSI where the each of the location give the distance that signal of LoRa could reach whilst the motion of vehicle is constant. From this experiment, certain level of RSSI value will be chosen to set as minimum boundary switcher for handover process state indicator and be used handover process algorithm in network switcher.

Ground Penetrating Radar (GPR) is widely used for non-destructive investigation of the shallow subsurface exploration especially in locating the buried infrastructure such as pipes, cables and road inspections. However, the interpreting hyperbolic signature of buried object in GPR images remains a challenging task. Therefore, this paper presented the classification of different materials based on GPR images using artificial neural network (ANN). In this research, GPR images so called the B-scan radargram represented by hyperbolic signature are firstly acquired and pre-processed. Then, the hyperbolic signature features are extracted using statistical techniques. The extracted features is then fed up as input to the multilayer perceptron (MLP) neural network classifier. A series of experiments have been conducted based on extracted hyperbolic features of different materials and shapes. Based on the results, the proposed method in classifying different materials based on GPR images using neural network showed promising results.

This study examines the cross-correlation coefficient of Mechanomyography (MMG) signals recorded from Palmaris Longus (PL), Flexor Digitorum Superficialis (FDS), Extensor Digitorum (ED) and Extensor Carpi Ulnaris (ECU) of forearm muscles during isometric muscle action. Nine healthy participants using dominant hand (mean±SD: age=24.78±1.79 year) volunteered in this study to perform submaximal to maximal (25%, 50%, 75% and 100%) isometric muscle actions with maximal voluntary contraction force of hand exercise gestures (grip strength supinated, grip strength pronated, finger flexion and pinch grip). During each isometric contraction, four-separated VMG sensor (TSD250A)-sensitive accelerometer type was used to record MMG signals on flexor and extensor side of forearm muscles. Maximum crosscorrelation coefficients at zero time lags were analyzed between flexor (PL & FDS) and extensor (ED & ECU) sides for each subject and force level to determine which sides show high level of association during isometric muscle action of forearm with different hand exercise gestures. The results showed maximum cross-correlation coefficients that range from R_xy = 0.801 – 0.943 for all hand exercise gestures performed. In addition, from the results obtained between flexor and extensor side indicates that the maximum range of cross-correlation coefficient for extensor side muscle groups were slightly higher than the flexor side muscle groups.

This paper discusses the prediction of movement from the vehicles with Long Range Wide Area Network (LoRaWAN) handover communication, based on Received Signal Strength Indicator (RSSI) algorithm. The experiment consists of observation on three different environment area which is rural, sub-urban and urban at road in Kedah and Perlis state in Malaysia. This three-environment area measured based on the past studies. Handover session with a moving vehicle, predicted the movement based on direction of Global Positioning System (GPS) latitude and longitude. The communication session transferring from previous cell area network that pass by the vehicle to a new targeting cell area network with the base station connection. From this experiment and study, it is shown that the predictable or movement from one cell to another cell based on RSSI algorithm with base station connectivity can be achieved.

Three phase induction motors are the workhorse of industry. Variable speed operation of the motor enables substantial energy savings through implementation of variable frequency drives. This project concerns on the design and implementation of three-phase voltage source inverter (VSI) for variable frequency drive. The focus was to generate variable frequency output suitable to be fed to the induction motor for the purpose of variable speed control. SPWM signal was generated through SIMULINK algorithm. The VSI circuit was designed, fed with SPWM trigger signal and simulated for variable frequency output in terms of line-to-neutral and line-to-line voltage across R-L load. In the experimental setup, the real VSI system was developed, the SPWM trigger signal generated through SIMULINK and interfaced to hardware using Arduino microcontroller. The output is dropped across resistive load and parameters line-to-neutral and line-to-line voltage was measured. Three set of output voltage with frequency of generated at 20Hz, 50Hz and 70Hz. Thus, experimental result validated that variation of frequency of voltage output of the VSI has been successfully achieved.

Bone implant is a promising technique to heal the fracture site where it focusing to stabilize and support the bone. Improvement in bone implant are required in order to ensure better healing process because some failures may happen due to the loss of implant stability. Previous studies showed there are lack of investigation on bone implant. Early detection of screw implant loosening is crucial for faster repairing and a healthy ready-to-operate. The purpose of this study is to investigate the signal strength of electromagnetic field induction on human bone by employing a single channel magnetic induction spectroscopy using flat spiral coil sensor. This study also to evaluate the various type of coil sensitivity of the magnetic induction spectroscopy on femur bone. The electromagnetic field induction will be analyse using the COMSOL Multiphysics simulation software to collect the data of received signal strength. The promising experimental results suggested the feasibility of the proposed electromagnetic induction spectroscopy.

Gas Distribution Mapping (GDM) can be defined as visualization for the spatial distribution of gas concentration in an unknown indoor environment. Recently, incidents of gas leakage have increased gradually worldwide. This will significantly increase the casualty rate at the workplace. This paper focusses on the implementation of Kernel DM based SLAM- GDM on Robot Operating System (ROS), a novel approach that will be useful in the effort of localizing the gas source during gas leakage inspection. This approach has been conducted successfully by using ROS-architecture implemented system, tested in simulation and on a mobile robot. The result shows that this implementation achieved average accuracy of 0.71m and 0.32m and therefore deemed suitable for gas source location prediction.

Mycobacterium Tuberculosis (MTB) is a common airborne infectious disease that leads to millions of deaths every year worldwide. It is still one of the top ten causes of death and the victims of TB are more than HIV/AIDS in 2017. Traditional approaches for MTB detection are either take a long time, unreliable or high cost. The electrochemical techniques (ECTs) as improved and inexpensive approaches to detect the MTB. Many of ECTs were used in MTB detection such as differential pulse voltammetric (DPV), cyclic voltammetric (CV), square wave voltammetric (SWV), amperometric and impedimetric. Principle of the MTB detection using ECTs depends on DNA hybridization of the MTB on the working electrode of ECTs. The researchers developed biosensors or aptasensors and used them for ECTs analyzing to detect the MTB. They developed various biosensors from various composite and DNA probes but all the developed composite of the biosensors were used to coat the electrodes that used in ECTs. Many types of electrodes and electrolytes were used in MTB detection. The most used ECTs in MTB detection is DPV and CV while the least used is amperometric. The ECTs for MTB detection achieved high sensitivity, reliability, low detection time and very low detection limits.

Electrocardiogram (ECG) is undoubtedly the gold method for assessing heart activities. However, the ECG machine for measuring ECG is mostly only available in health care center especially in the third-world countries. Therefore, this research proposed to utilize the product of National Instruments i.e. LabVIEW Biomedical Toolkit and data acquisition device as an alternative approach to acquire ECG features and heart beat rate (HBR). As for performances verification, the NI-LabVIEW instruments are tested simultaneously with two types of ECG machine which are Philips PageWriter TC 30 and KenZ ECG 108. The results indicated that the ECG features and HBR obtained from NI-LABVIEW instruments are agreed well with the measurements attained from both of the ECG machines. Hence, this study proved that NI-LabVIEW instruments are highly potent for measuring ECG features and HBR. Moreover, the computerized NI-LabVIEW instruments with the self-interpretation function can always outperform most of the conventional ECG machines that are still required cardiologists for manually interpreting the results printed on the ECG paper.

To evaluate the various internal quality attributes of fruits, NIR spectroscopy techniques are undoubtedly quick and non-destructive tools. Acidity (pH) is one of the main quality attributes in mango fruits. Generally, it is being predicted in destructive way. The aim of this research was to develop calibration model and prediction of pH in Perlis Sunshine mangoes using NIR spectrometer. The transmission spectra of Sunshine mangoes were acquired in the wavelength range from 300 to 1000 nm. The effects of different types of pre-processing methods and spectra treatments, such as baseline correction, multiplicative scatter correction (MSC), Savitzky-Golay (SG) smoothing, second order derivative (SG) and normalisation were analyzed. The prediction models were developed by partial least squares (PLS) regression. The coefficient of determination (R²) of pH was 0.928 and the standard error of cross-validation (SECV) was 0.153. The results indicated that by using the NIR measurement system, in the suitable spectral range, it is possible to predict the pH of mango fruits by non-destructively.

Wheelchair is very important as a means of transport by a person who is unable to walk as a result of illness or injury. On the other hand, smartphones have become common device that allow users to remotely control many electrical and electronic appliances. This paper presents a study on the mobility performance of wheelchair using smartphone accelerometer. The powered wheelchair is presented by a two-wheeled robot car which emphasized on the control layout platform for the smartphone's tilting directions with suitable tolerance values. A basic motion's algorithms for the movement of robot car have been developed, and its mobility performances on the cornering and straight-line direction are analysed.

Energy scavenging has become a topic of interest toward green and environmentally friendly. Vortex generated in the wake region behind a bluff body is potential to be harvested and transform the mechanical vortices into electrical energy. Therefore, the study of the fluid flow and piezoelectric material is highly potent and a promising research area. Thus, the vortex-induced vibration (VIV) energy harvesting of two bluff bodies in a tandem arrangement is investigated using the 3D simulation. There are four sets of bluff bodies with various distance ratio, x/D from 3 to 6, were analyzed in this simulation. The coupling analysis was analyzed using ANSYS Transient Structural and computational fluid dynamic Fluent module. The triangular width, to 100 mm and the fluid average flow velocity is 1.46 ms^−1, which corresponds to Reynold number (Re) of 10k. Four types of analysis have, out; velocity vector, turbulence kinetic energy, piezoelectric tip deformation, and generated output voltage. The results show the bluff shape with spacing ratio 6 has better vortex shedding formation and turbulence kinetic energy formed on both upstream and downstream triangular bluff body. The tips mesh displacement for piezoelectric is highest at space ratio 6 with 7.0 x 10⁻⁵ essential, and after 10s the beam deflection is continuing at around 1 x 10⁻⁵ m. While the voltage generated is, at spacing ratio 6 with 0.2 v produced. The obtained results are essential to VIV analysis of the dual triangular cylinder with piezoelectric film, which is vital in energy scavenging knowledge.

Tree detection and delineation has become one of the important factors that contribute to precision agriculture specifically in plantation industry and efficient forestry management. However, this requires tools and technology that would give reliable information and high accuracy data processing. Recent researches aimed at providing this goal by utilizing the advancement of available remote sensing technology and integrate various algorithms. This article reviews these researches with a focus on algorithms applied to remote-sensing imagery for the purpose of tree detection and delineation. It categorizes and evaluates those methods with the respective types of remote sensing imagery used in the area to evaluate the influence of these factors on the method. Finally, it summarizes the finding of the current algorithms, and suggests on the new development that can be expected in the future.

Electrical resistance tomography (ERT) reconstructs internal resistance images of the field from electrical measurements on the surface. The objective of this paper is to obtain the tomogram of the liquid-gas regime of ERT system by using multiple excitation sources. A linear back projection algorithm was implemented as a basic algorithm. Then, it was modified to suit the implementation of multiple excitation sources. A liquid-gas regime in the steel pipe was used as a tested medium. The analysis and performance of the image reconstructed which applies multiple excitation sources were compared with the single excitation technique. As a common strategy applied a single excitation source and produce problem especially at the center of the pipe, this paper mainly focuses on image reconstruction of multiple excitation sources. As a conclusion, 50% of the excitation source cannot produce the image as expected. Tomograms that produced by using the single excitation source were much better than the images obtained by using 50% of the excitation source.

This study presents the comparison between two meta-heuristic algorithms for curve reconstruction of Arabic fonts. The two algorithms namely Sine Cosine Algorithm (SCA) and Whale Optimization Algorithm (WOA) are used for comparison purposes. These algorithms have been applied to optimize the solution of curve fitting problem by using cubic Bezier curve. There are several pre-processing steps involve before curve fitting process has been done. Firstly, extract the boundary. Then, detect the corner of outline fonts and the third step is chord length parameterization. The aim of this study is to find the best curve that fit the data. Sum Square Errors (SSE) has been used to calculate the error given by original image and parametric curve. The finding of this study shows that, WOA gives better curve approximation based on the errors result than SCA. This algorithm significantly improves the shape of the corner of the outline Arabic fonts.

This paper presents the development of an intelligent control of a modular power assisted stretcher movement on flat and sloped floors for indoor environment. Conceptual design techniques are used to select the best concept for prototyping power assist systems. The operator's movement intention and related voltage data is sensed by an Eight-axis hall effect sensor. It proposes the use of PID controller to optimize and maintain the velocity of the stretcher adapted to different weight and floor conditions. A stepping motor is applied to drive each wheel to respond appropriately to different floor and steering conditions (flat or sloped, straight section or turns). Without assistance it would be impractical for an operator to move a loaded stretcher (with patient and accessories) weighing up to approximately 100 kg on 15 degrees straight slope-up / slope-down floor sections. The power assist application turns this into an easy operation.

Ground penetrating radar (GPR) is a non-destructive tool for subsurface investigation becomes an emerging trend in many applications especially in detecting infrastructure utilities. However, interpreting of GPR data in terms of target's geometry is remained a challenging task due to the existence intra/inter-class variations and subtle changes of hyperbolic signature pattern in GPR data. To addressed this issue, this paper proposes to classify the basic shape (i.e. cubic, cylindrical and disc) of buried object using multilayer perceptron neural network. The GPR raw data firstly is pre-processed based on A-scan and B- scan GPR data. Then, the combined statistical features with gray level co-occurrence matrix (GLCM) and Hu moment invariants are extracted from A-scan and B-scan. The extracted features are then fed as input to multilayer perceptron to classify the shapes (cubic, cylindrical and disc) of buried object. Based on the experiment conducted, the highest recognition rate has been achieved is about 83.3%. Thus, the proposed method of using combined statistical with GLCM and Hu moment invariants features to classify shape of buried object using multilayer perceptron is promising.

This paper presents the development of gas sensor module consisting of metal oxide (MOX) gas sensors for indoor air contaminant measurement. The first phase of the studies involves the design of PCBs board for six different metal oxide (MOX) gas sensors particularly CCS803, MiCS5524, GM-402B, GM-502B, GM-702B and MiCS6814. Next, the main board consisting of temperature, humidity and another two MOX gas sensors (i.e TGS2600 and TGS2602) was developed for the acquisition of the sensors. A partially closed chamber was designed and fabricated to fit the main board and allows inflow and outflow of gases. The responses of the gas sensors were measured in a closed room, with the air conditioner turned on and off periodically. The results show that all the MOX gas sensor used are affected by temperature and humidity of the environment. Therefore, the sensor response drift needs to be corrected in order to obtain reliable indoor air contaminant measurement.

Simultaneous localization and mapping (SLAM) is considered to be one of the primary tasks for autonomous robot navigation process. The first element required by the robot to start the navigation process is to create and assemble an accurate map of the surroundings or ground plane to envision its location. Mobile robots that use 1D or 2D sensors such as laser range finder, ultrasonic or sonar for the mapping process cannot provide enough information on the obstacle locations while 3D sensors like Kinect can. However, if the 3D sensor is not mounted correctly, the orientation of the 3D image obtained will also be affected and thus producing inaccurate maps. In this project, Microsoft Kinect was used to scan the environment. RANSAC algorithm was implemented to detect the ground plane. The orientation of the ground plane was corrected, and the depth data acquired by Kinect was then converted into 2D map. It was found that the methods applied have successfully mapped the obstacles detected.

This paper presents the used of histogram of oriented gradient (HOG) for facial expression recognition using support vector machine (SVM). In this work, the facial expression images are firstly preprocessed by face detection and cropped images. Then, HOG method is adopted as feature extraction on facial image. The ability of HOG to preserve the local information and orientation density distribution in facial images suitable as shape descriptor for facial expression. It divides the image into cell or patch that has magnitude and orientations. The extracted HOG was then concatenated into histogram bin to form one feature vector before feed into SVM classifier. Both JAFFE and KDEF datasets were employed to evaluate the performance of proposed method. Based on results, the average recognition rates of JAFFE and KDEF datasets are 76.19% and 80.95% respectively. The results show that the performance of expression surprise has outperformed compared to others expression while expression fear contributes the lowest recognition rate. Thus, utilization of HOG features with SVM classifier have shown the promising results in recognizing facial expression.

The massive deployment of wireless infrastructure especially in the trending 4th Industrial Revolution (IR4.0) has opened the opportunity for wireless-based positioning and localization. The works on wireless positioning system revolves on two techniques, the geometrical technique employing either triangulation or trilateration, and mapping technique employing signal fingerprint method. The later is more favoured since the accuracy was superior to geometrical technique. However, the signal fingerprint method required enormous efforts in order to record and stored the signal fingerprint. This paper investigates the new approach in constructing the signal fingerprint using organic approach. The signal is recorded using the developed apps, and then the user sends the signal data into the cloud. The organic fingerprinting data were compared to the conventional fingerprinting. The preliminary results show promising accuracy in the experimented indoor environment.

Oil palm is known as one of the most important resources in Malaysia which able to be exported to other countries. Oil palm, also known as "Elaeis guineensis", has many usages and benefits. To ensure the productivity and quality of the oil palm, the health status of the tree needs to be monitored, as they can be prone to various diseases. This paper is going to discuss current methods other researchers have adapted that able to be used in monitoring oil palm trees health. In addition, some analyses and conclusion were drawn to address related issues in the area.

This work focuses on comparing gas source localization accuracy of three types of gas source localization methods which are the conventional method, weighted-averaging mean method and variance map from DM+V method. Gas source localization accuracy in this work means the distance error between the calculated source location and the actual source location. All methods were implemented offline using the measurements collected from real-time gas source localization experiments that were done beforehand. The dataset consists of gas sensor measurements, and mobile robot's location in a 3m by 6m test area. The goal of the work is to shine a light on why recent works in the field are more focusing on improving probabilistic and map-based algorithms. This work also shows how the kernel size of kernel-based gas distribution mapping might affect the gas source localization accuracy in a small indoor area.

A study was conducted to investigate whether reflectance data of hyperspectral image of an area could be used to extract related physical features to produce mapping of vegetation density. This paper explains on estimating percentage of vegetation coverage based on Normalized Difference Vegetation Index (NDVI). Image segmentation based on thresholding was used to separate different features of the land entities like soil, water and road. From here, NDVI values can be integrated for further segmenting the vegetation features. The colour segmentation method is then able to classify the vegetation according to their density level, which can be used to determine tree crown density. Test conducted towards a hyperspectral image shows that different density level can be extracted, where it contains about 7.5% high level tree crown density, 1.8% medium crown density and 5% low. More tests need to be conducted in order to proof the workability of the developed algorithm in analysing hyperspectral images from tree crown density mapping.

This paper proposed a RF-based mapping-based indoor localization technique to locate their packages and items. The Received Signal Strength Indicator (RSSI) is commonly used in trilateration to allocate the packages or items location. However, several problems need to overcome to increase the accuracy of the RSSI signal propagation for the localization. The problems include the complexity of the indoor environment will affect the radio waves experienced multipath effect such as the signal may be reflected, refracted, absorbed or blocked by other matter. The paper is designed to study the propagation of radio waves in indoor conditions, identify the type of antenna used and determine the accuracy of the localization model. The RSSI is recorded for every 1m until reached the maximum read range of 20m to study the behaviour of radio waves propagation and the relationships between RSSI and distance. The RSSI mapping experiment was carried out to determine the type of antenna used to carry out the experiment. The localization is performed with mapping-based technique to predict the random coordinates in the testbed. The predicted coordinates and actual coordinates were compared to determine the accuracy of the localization model. The error between the actual coordinates and predicted coordinates can achieve an average of 0.95 meter. The RF mapping-based indoor localization technique results shown that the effectiveness of the techniques to be implemented in a company to improve the efficiency in the production line.

Mapping is one of the elemental application of the mobile robot. The map is created using the mobile robot by employing sensors such as camera, sonar and laser sensor. One of the most popular mapping methods is the Simultaneous Localization and Mapping (SLAM). SLAM allows the map to be created while localizing the robot location in the map at the same time. GMapping is one of the widely used algorithms in SLAM which will be used in this project. The mobile robot is equipped with a Hokuyo Laser Range Finder sensor and netbook. The router is used for wireless communication between the mobile robot and the user. The GMapping is done in two different locations of different lab size and amount of features in the area. Three trial is conducted to investigate the effects of different parameters such as robot speed, mapping delay and particle filter on the mapping quality. The results show a significant difference in terms of mapping accuracy and the time taken to complete the process as the parameter changed from the three trial. As a result, the parameter used in the second trial, robot speed 0.1333m/s, mapping delay 1s and particle filter 30 is considered as the best based on the time taken and the map accuracy.

This document discusses the novel approach to localize human location based on the current zone via k-mean clustering. A pilot experimental analysis of k-mean clustering to group similar RSSI pattern is compared to user defined zones. The dataset collected has demonstrated the difficulties of deploying trilateration and fingerprinting methods in dynamic conditions, as it introduces fluctuating RSSI measurements. The k-mean clustering was proposed to validate antenna placement to divide the testbed into zones, and to create a baseline accuracy which can be compared with other algorithms in the future. It was found that, for Zones 2, 3 and 4, the k-mean clustering of antennas agree with the planned antenna placement grouping. However, for Zone 1, there are differences, which may be attributed to a large metal obstacle occupying the zone. The k-mean clustering also recorded a peak accuracy of 79% with k=4, which agrees with the number of planned zones.

Small spacecraft requires capable processors with energy efficiency, low cost and low computational burden while maintaining the output tracking accuracy. This paper presents the extension of work in [1], to enhance the transient performance using particle swarm optimization (PSO) on decaying boundary layer and switching function thorough error feedback (DBLSF) in Sliding Mode Control (SMC). Generally, SMC is known for having chattering as the main drawback which can introduce wear and tear to moving mechanical parts. As a solution, a DBLSF proposed in [1] and capable of eliminating the chattering in SMC while considering the essential requirements for small spacecraft operation. Then, the extension implemented on spacecraft's attitude, which is one-of-six subsystems in spacecraft, used to orient the spacecraft referred to reference objects and control the dynamics of a spacecraft time-to-time according to the needs. However, the SMC's transient response can be tuned using some coefficients in the SMC algorithm. The parameters in [1] were tuned using outputs observation technique. In this paper, then, an improvement is introduced to optimize the outputs by adding a PSO in the SMC-DBLSF in term of transient performances and accuracy while reducing the chattering permanently.

Human identification is a very important subject in computer field and has been researched widely. This paper proposes human identification system based on Kinect's individual sensor as well as combination of all its available sensors. In the first part of the project, each sensor on the Kinect (i.e IR depth sensor, RGB camera and microphones) was used for skeleton recognition, face recognition and speech recognition respectively. Then, these individual recognition methods are combined as step-by-step process into a multi-sensor recognition system. Few experiments were carried out to test the reliability of the developed human identification systems. The results show that multi-sensor based human identification system is highly efficient compared to the single-sensor system. This is because, multi-sensor recognition system involves many recognition stages as each recognition stage needs particular biometric information of the user.

Omnidirectional mobility plays an important role for mobile robots to provide flexibility in their motion. However, most of the existing omnidirectional mechanisms are only superior in indoor applications, but inappropriate for outdoor purposes. This paper introduces the design and development of an omnidirectional mobile robot with reconfigurable screw-wheeled mechanism. The orientation of the screw wheels is adjustable in order to enable flexibility in traction control to provide more propulsion drive in certain conditions. The experimental results are given to demonstrate the performance of the omnidirectional mobile robot on different ground surfaces and different wheel's orientation.

Hazardous gas in the confined space is an issue that often causes danger to the worker while doing the tasks i.e. installation, cleaning and maintenance. The nature of the area may exposure the worker to hazardous gas leakage. It is important to monitor the hazardous gas in the surrounding area before the worker start their task inside the confined space. This paper proposes a study of developing a web-based mobile olfaction system for confined space hazardous gas monitoring. The system is integration between a mobile robot and an electronic nose (e-nose). The system will reduce the danger faced by the worker from the hazardous gas. The e-nose consists of a sensor array, embedded controller, electronic components and wireless GSM communication. The instrument sensors i.e. hydrogen sulphide, oxygen, carbon monoxide and methane are used to acquire the confined space air sample. It will also acquire the environment temperature and humidity. The robot which piggyback the instrument will manoeuvre through the environment while acquires data. The data acquired were transmitted and monitored wirelessly to the web-based information system. The website layout is designed using HTML and JavaScript. Data received from e-nose will be stored into a database, MySQL server before being process by the website. The data will go through the conversion process from digital value to PPM. Then the data will be classified using ANN to obtain the Hazardous Gas Concentration Index (HGCI). The website will display the HGCI mapping of the area in real-time. Results show that the system performance is good and can be used to monitor the hazardous gas in the confined space.

Malaysia is one of the largest palm oil producers in the world. To meet the production, its plantation needs to be monitored accordingly. Manual monitoring is no longer feasible due to energy, cost and time consumption. Remote sensing technology such as satellite images can be utilized for palm oil monitoring as they are cost effective, time and energy saving, and able to enhance the possibility of classifying the vegetation through spectral and texture analyses. Due to its impact towards Malaysia economy, an effective monitoring of palm oil plantation can be done by using satellite image to maintain the sustainability of palm oil in Malaysia. This project utilized the computation of Normalized Difference Vegetation Index (NDVI) and Soil Adjusted Vegetation Index (SAVI) to determine the plant health. Relationship of NDVI and SAVI with respect to temperature and precipitation is studied based on the data obtained from Malaysian Meteorology Department. By using regression model technique, the relationship of the plant health based on NDVI analysis, nitrogen content based on SAVI analysis and temperature is found out to be highly correlated while plant health and precipitation is found out to be no or low correlated. The results can be utilized by the plantation operators, especially independent smallholders in monitoring the health status of the palm oil trees to ensure high productivity.

Ergonomics knowledge is crucial in any industries, without knowledge applied, one easily exposed towards the risks of the working environment. Based on the survey conducted showed awkward posture and heavy lifting were the most highlighted issues occurred. Aim of this study is to analyse the working posture in mechanical assembly manufacturing industry and analyse the effect of the improvement workstation. The use of Rapid Upper Limb Assessment (RULA) in order to analyse the most level of risk of workers during working, develop human manikin following the anthropometric data collected from industry real population, as well to replicate the posture of the workers, simulate the working process via simulation using DELMIA V5. Results reported one of the highest ergonomic risk occurred in this department is drilling activity. The manikin successfully developed with 105 measurements for 685 male workers. Simulations effectively simulated using DELMIA V5 in virtual environment. Step ladder suggested as virtual and this improved the value of RULA score. In conclusion, the software and method used were powerful tool in designing the virtual improvement and this research expected to create a design guideline for the industry that help workers with an ergonomic way at the workstation before and during working.

Human following robot is very common in this technology era. Human following is a technique used by robot and autonomous vehicles to follow a human within a specific range. In this case, communication between the human and the robot is the most significant factor where sensor is needed to ensure its successfulness. This paper discussed a human following robot system that utilized ultrasonic sensor. Ultrasonic sensor is preferred for human following robot due to its wide detection area, less light dependency, the ability to detect glass and shining wall, smaller in size, lightweight, use a very low memory, cheaper than Laser Range Finder (LRF) or camera and lower power consumption. However, being a specular type of sensor and not narrowly focused, this can cause wrong estimation and recognition of human, especially at the legs. Therefore, necessary accompanying algorithms should be developed to encounter this issue. Here, three pieces 40 kHz ultrasonic sensors are mounted on the four-wheeled mobile robot platform. Ultrasonic sensors used to detect the human's leg, and the data collected from the sensors were used upon the design of sensor platform. In order to detect human's leg, the sensor platform was tested repeatedly to obtain optimum data. The data which have been collected from the ultrasonic sensor will then be interfaced with Arduino software. Thus, this paper outlines a set of benchmarks for human following and briefly evaluates its performances. The human following robot was tested in a real laboratory environment. The results were good, where the implementation of this algorithm is able to produce an accurate decision.

Ground penetrating radar is one of the promising non-destructive investigation for shallow subsurface exploration in locating buried utilities. However, interpreting hyperbolic signature of buried objects in GPR images remains a challenging task since the GPR signals are easily corrupted by environmental noise and cause misinterpretation of the size and geometry of subsurface object from the GPR raw profile. Therefore, this paper proposes Discrete Wavelet Transform (DWT) and principal component analysis (PCA) to classify geometry of buried object using k-nearest neighbour. (k-NN). The GPR images firstly being pre-processed. Then, the GPR images are decomposed using DWT into four sub-bands which are LL (Low-Low), LH (Low-High), HL (High-Low) and HH (High-High). The sub-bands LL or a coarse approximation coefficients was extracted as DWT features in order to classify the shape of buried objects. Since DWT features do contain high dimensional data, thus PCA is used to reduce the dimensional features from higher to lower space by linear transformation. The new projected features were then classify using k-NN classifier into four shapes which is cubic, cylinder, disc and sphere. A series of experiments have been conducted on extracted DWT and PCA features from hyperbolic signature of buried objects having different shapes. Based on the results, the proposed method had achieved the average recognition rate of 99.41%.

Outdoor localisation can offer great capabilities in security and perimeter surveillance applications. The localisation of people become more challenges when involving with the nonlinear environment. GPS and CCTV are two localisation techniques usually use to localise human in an outdoor environment. However, they have weaknesses which result in low localisation accuracy. Therefore, the application of Device-free localisation (DFL), together with the Internet of things (IoT) is more appropriate due to their capability to detect the human body in all environmental conditions, and there is no problem losing signals as faced by GPS. This system offers excellent potential in humans localisation because humans can be detected wirelessly without any tracking device attached. In developing the DFL system, the main concern is the localisation accuracy. Although the existing DFL system gives significant result to the localisation, the accuracy is still low due to the large variation in RSSI values. Hence, a Radio Tomographic Imaging-based ANN classification (RTI-ANN) approach is proposed to increase the localisation accuracy. This Artificial Neural Network (ANN) is designed to learn the Radio Tomography imaging (RTI) input for classification purpose. Even though the RTI gives a good result to the localisation, however, it suffers from smearing effect. To eliminates this smearing area and background noise, pre-processing of the RTI image is required. Thus, extracting the valuable information technique from the RTI image has been proposed. By extracting the valuable information data from the RTI image, about 61% to 66% of the smearing noise is removed depending on the size of the RTI image. Only data directly associated with human attenuation used for training and learning of ANN. The experimental results show ANN system can localise human in the right zone for a given dataset.

Caused by the energy revolution, energy saving becomes a bigger topic than ever. This paper is about the developing of a new controlling system for air conditioning, which could save up to 66% of the energy consumed.

The implementation of intelligent transportation systems, or what is otherwise referred to as IoT and big data analytics in transportation system are affected by several factors. Considering the overall impact of a definite and comprehensive plan that involves business, infrastructure, and organizational administration, improved customers satisfaction can better be achieved when all these antecedent factors are investigated, and their elements are determined. There is no recorded study on multidimensional implementation model where more than one of them is considered. This paper, therefore, stems from the general background of implementation success factors but aims to discover the generality of the definite plan and procedure that must be executed for a successful implementation of big data analytics and IoT-oriented transportation system. This paper employs a systematic literature review and an in-depth interview to address the obvious limitation. This paper also discussed the qualitative approach that was used to gather data on the successful implementation factors in the transportation system. The deliverable from this research will be the implementation success factors for the big data analytics and IoT-oriented transportation system.

This paper presents a workspace design of a tailor-made learning environment for students using the product-service systems (PSS) concept. Existing research regarding workspace design focuses on the design of the interior of classrooms to ensure that the learning process proceeds with minimum stress and maximum effectiveness. None of the current research had discussed the workspace outside of the classroom. Therefore, the main objective of this study is to propose a learning environment workspace in which students can gather, study, relax, socialize, eat, and work during class gaps. For this purpose, a questionnaire was distributed to identify the behavior of the users to generate a list of requirements in the development of the learning workspace. This paper focused on students at the Engineering Faculty and the learning environment workspace was called the OneForAll Center. Out of 10 functions of the proposed OneForAll Center, only four functions had the highest priority: a space to use a laptop and internet connection, a television, a vending machine and a printing section. The OneForAll Center has been illustrated in a 3D model and shown to the students, in which most of them agreed with the proposed services and facilities such as a space to use the laptop and internet connection and a printing section with a printing service. The OneForAll Center has been identified as a solution to improve learning outside the classroom, specifically to the provision of services to students.

This paper presents the development of dynamometer to measure cutting force in turning operation in particular the tangential force (F_t). The dynamometer utilises a straight bar load cell that connected to holder tool to measure at least one axis of force. The designed dynamometer is capable of computing the forces acting on the workpiece in turning operation using any data acquisition system. Two materials of low elastic modulus, Polyoxymethylene (POM) and Polytetrafluoroethylene (PTFE), were employed as workpieces to test the performance of dynamometer. The performance of the dynamometer to measure one axis of cutting forces is considered satisfactory mainly because of the simplicity and low cost are the main requirements in design and construction.

Deficient soil fertility is a major constraint in crop productivity. Soil characteristics such as pH, bulk density, electrical conductivity, moisture content, etc. play a major role in the health and yield of crops. Major nutrients such as Nitrogen (N), Phosphorus (P) and Potassium (K), are essential for the growth and yield of the crops as well. On account of the importance of soil physical properties and NPK concentration on the soil dynamics and plant development respectively, the study area was spatial distributed and geo-statistically analyzed to produce contour maps that can be used by decision makers in farm management for increased crop productivity.

Connected car has become one of emerging technology in the automotive industries and vehicular communication studies on the other hand, has become one heavily researched topic in academia which focuses on the performance of various parameters upon vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication-based in the IEEE 802.11p vehicular communication studies. In urban environment, obstructions from existence of building, trees, street's layout and other obstruction or objects can strongly influence the receive signal levels at 2.4 GHz frequency band. To ensure reliable and stable communication link for efficient future deployments of cooperative RSUs deployments, will require analyses on the signal strength upon environments that generally characterized by challenging propagation conditions. This article presents the results of experimental field testing of signal characteristics of different types of road topology such as grass, gravel and tar surface on open field environment. The results show that road topology or streets layout have an effect on V2I communication and the attenuation resulted from certain types of road's surface can give significant impact on signal propagation.

The moisture distribution in the silos depends upon various seeds parameters such as type and size of seeds, amount of storage, external weather, and storage period as well as structural and environmental factors. It is very difficult to predict moisture distribution in silos effectively while taking all the above aspects into consideration. This study aims to investigate the efficiency and accuracy of localization of moisture distributions sensing in agricultural silo. The work is mainly focussed on three major elements: Radio Frequency (RF), tomographic imaging and classification process using machine learning. In particular, RF-based signal and volume tomographic images are used to predict the moisture distribution. Furthermore, computational intelligence techniques such as artificial neural network (ANN) is applied to develop models based on previous data. The generalization of these models towards new set of data is discussed in making sure a successful application of a model. A detailed study of the relative performance of computational intelligence techniques has been carried out based on different statistical performance criteria.

The number of cases of Breast cancer seems to be constantly increasing worldwide in the recent years. According to 2012 world cancer report, the breast cancer incidence rate in Asia appears to be 37.5 and the mortality rate is 13.21. As per the recent statistics report of 2018, it has been estimated that the number of new cases of breast cancer is 14% in India. The percentage is higher when compared to lung cancer, oral cancer and other types of cancers. It has been identified that it occurs mostly in the age group between 41-50 years and the distribution of this is found to be 42%, whereas it is 18%, 24% for the age groups of 31-40 and 51-60 respectively. The five stages of survival include the survival rate of 100%, 98%, 88%, 52%, 16% for each stage respectively. Diagnosis in the early stage can reduce the death rate. Even during the stage 3 if the cancer is predicted, it can be treated. Machine learning using Artificial Neural Network(ANN) techniques can be effectively utilised for the prediction of cancer. In the proposed method, networks designed using various ANN Algorithm are used to predict the breast cancer if it is benign or malignant. Three different networks such as feedforward back prop, Cascade forward and layer recurrent has been implemented. The performance measures such as Accuracy, error, specificity, sensitivity, positive predictive value and negative predictive value are obtained. From the table 2, it is inferred that among these algorithms, feed-forward algorithm has better performance compared to other two algorithms.ANN has been a powerful tool for analyzing the data when there are non-linear interactions between the input and the output to be predicted. The results show that the accuracy of ANN for the prediction was better than other approaches. The Wisconsin breast cancer database has been used for the analysis. The results are very competitive and can be used for diagnosis, prognosis, and treatment.