Search

Vehicles with advanced active safety technology can decrease the significant traffic accidents that can lead to death. This active safety frontier falls under primary safety in the European New Car Assessment Program (Euro NCAP) 2025 Roadmap, which has become one of the overall safety rating initiatives toward safer vehicles. Some frontier active safety technologies will be assessed, including autonomous emergency steering (AES) and autonomous emergency braking (AEB). However, the New Car Assessment Program in Southeast Asian Countries (ASEAN NCAP) only focuses on AEB technologies. Hence, this work discusses the existing papers on AES assessment, AES demand, AES control, AES system with Artificial intelligence, and AES testing methodology. Three articles from the industry discussing the AES function in passenger automobiles were found as a result of an article search using the Google search platform. Other terminologies like emergency steering control and emergency steering assist are used instead of AES. However, the principle remains the same. The three categories have been recognized from all of the document results: road adhesion condition, driver condition identification, and rear-end collision. However, only the rear-end collision situations are further investigated in this work to recognize the currently available approach used by previous studies. According to the review findings, just a few AEB intervention systems are now accessible, while AES technology is still in its early phases. That might explain the lack of exact evaluations and effective remedies. As a result, this research aims to offer evidence supporting the proposed methodology for assessing and evaluating AES in the ASEAN NCAP rating scheme. Besides that, this study can also help industries such as automakers and automotive vendors leverage the guidelines to fit the AES in their future models.

The development of ankle foot orthoses (AFO) for lower limb rehabilitation have received significant attention over the past decades. Recently, passive AFO equipped with magnetorheological brake had been developed based on ankle angle and electromyography (EMG) signals. Nonetheless, the EMG signals were categorized in stance and swing phases through visual observation as the signals are stochastic. Therefore, this study aims to classify the pattern of EMG signals during stance and swing phases. Seven-time domains features will be extracted and fed into artificial neural network (ANN) as a classifier. Two different training algorithms of ANN namely Levenberg-Marquardt (LM) and Scaled Conjugate Gradient (SCG) will be applied. As number of inputs will affect the classification performance of ANN, different number of input features will be employed. In this study, three participants were recruited and walk on the treadmills for 60 seconds by constant the speed. The ANN model was designed with 2, 10, 12, and 14 inputs features with LM and SCG training algorithms. Then, the ANN was trained ten times and the performances of each inputs features were measured using classification rate of training, testing, validation and overall. This study found that all the inputs with LM training algorithm gained more than 2% average classification rate than SCG training algorithm. On the other hand, classification accuracy of 10, 12 and 14 inputs were 5% higher than 2 inputs. It can be concluded that LM training algorithm of ANN was performed better than SCG algorithm with at least 10 inputs.

A passive controlled ankle foot orthosis (PICAFO) used a passive actuator such as Magnetorheological (MR) brake to control the ankle stiffness. The PICAFO used two kinds of sensors, such as Electromyography (EMG) signal and ankle position (two inputs) to determine the amount of stiffness (one output) to be generated by the MR brake. As the overall weight and design of an orthotic device must be optimized, the sensor numbers on PICAFO wanted to be reduced. To do that, a machine learning approach was implemented to simplify the previous stiffness function. In this paper, Non-linear Autoregressive Exogeneous (NARX) neural network were used to generate the simplified function. A total of 2060 data were used to build the network with detail such as 1309 training data, 281 validation data, 281 testing data 1, and 189 testing data 2. Three training algorithms were used such as Levenberg-Marquardt, Bayesian Regularization, and Scaled Conjugate Gradient. The result shows that the function can be simplified into one input (ankle position) – one output (stiffness). Optimized result was shown by the NARX neural network with 15 hidden layers and trained using Bayesian Regularization with delay 2. In this case, the testing data shows R-value of 0.992 and MSE of 19.16.

This work presents a hybrid metaheuristic, namely a hybrid marine predator algorithm - hide object game optimization (MPA-HOGO). MPA-HOGO is built by combining two shortcoming metaheuristics: marine predator algorithm (MPA) and hide object game optimization (HOGO). The distinct strategy in MPA-HOGO is as follows. MPA-HOGO utilizes an exploration-to-exploitation strategy gradually as the iteration goes on. It uses the global best solution evaluated in every iteration to ensure that the final solution comes from the best solution found during the iteration. It also eliminates the normalized score of every solution to reduce the computational process. This metaheuristic is then tested by solving 23 classic functions. MPA-HOGO is benchmarked with particle swarm optimization (PSO), Komodo mlipir algorithm (KMA), MPA, and HOGO. The result indicates that MPA-HOGO is superior among these metaheuristics. MPA-HOGO is also competitive in solving practical optimization problems. Through simulation, MPA-HOGO produces 13%, 7%, and 2% lower total costs than PSO, MPA, and HOGO respectively in solving inventory management problems in the vendor-managed inventory system. [ABSTRACT FROM AUTHOR]

This work proposes a new simple metaheuristic optimization method inspired by the social forces model used in pedestrian dynamics. The proposed model is a swarm-based model where a collective intelligence is shared among the agents, consisting of several persons or agents who walk over the search space to find the best solution. Each time a person finds a better solution, they share it with another person. The proposed model is then evaluated by implementing it to solve ten benchmark functions. The five are multimodal functions (Ackley, Rastrigin, Griewank, Bukin, and Michalewicz) while the other are unimodal (Sphere, Rosenbrock, Bohachevsky, Zakharov, and Booth). The performance is compared with five metaheuristic algorithms: particle swarm optimization, darts game optimizer, shell game optimization, marine predator algorithm, and Komodo mlipir algorithm. The simulation result shows that the proposed method is competitive enough to solve multimodal and unimodal functions. The performance is superior in solving Michalewicz and Zakharov functions but less competitive in solving the Bukin function. The results also imply that there is no single algorithm that is the best in solving all kinds of problems. [ABSTRACT FROM AUTHOR]

One critical problem in Indonesia's national joint courses program, initiated by the ministry of education and culture of Indonesia, is the lecturer-course assignment problem. Although the lecturer-course assignment problem has been studied widely as part of the education timetabling problem, no existing lecturer-course assignment model suits this circumstance. The new cases in this program are as follows. First, this program is conducted online. Second, the participants are students and lecturers from different universities. Based on this problem, this work proposes a novel lecturer-course assignment model that suits this program. The lecturers' preferred courses and timeslots become hard constraints. The model has three objectives: (1) maximizing the educational quality, (2) maximizing the lecturers' time preference, and (3) minimizing the number of unserved classes. This model is developed by using integer linear programming and optimized by using cloud theory-based simulated annealing. This proposed model is then compared with the four previous lecturers-course assignment models. The first model concerns about minimizing the number of unserved classes, while the second model focuses on maximizing the education quality. The maximum number of classes per course for every lecturer is considered in the third model while balancing the lecturer’s load (teach, research, community service) is the feature of the fourth model. The research concludes that the proposed model is appropriate for lecture-course assignment in Indonesia’s national joint courses program compared to the previous models. Based on the simulation result, the proposed model performs moderately in education quality and several unserved classes. Meanwhile, the proposed model is the best in the timeslot preference aspect by creating a 25% to 28% higher total timeslot score than other previous models. [ABSTRACT FROM AUTHOR]

The development of ankle foot orthoses (AFO) for lower limb rehabilitation have received significant attention over the past decades. Recently, passive AFO equipped with magnetorheological brake had been developed based on ankle angle and electromyography (EMG) signals. Nonetheless, the EMG signals were categorized in stance and swing phases through visual observation as the signals are stochastic. Therefore, this study aims to classify the pattern of EMG signals during stance and swing phases. Seven-time domains features will be extracted and fed into artificial neural network (ANN) as a classifier. Two different training algorithms of ANN namely Levenberg-Marquardt (LM) and Scaled Conjugate Gradient (SCG) will be applied. As number of inputs will affect the classification performance of ANN, different number of input features will be employed. In this study, three participants were recruited and walk on the treadmills for 60 seconds by constant the speed. The ANN model was designed with 2, 10, 12, and 14 inputs features with LM and SCG training algorithms. Then, the ANN was trained ten times and the performances of each inputs features were measured using classification rate of training, testing, validation and overall. This study found that all the inputs with LM training algorithm gained more than 2% average classification rate than SCG training algorithm. On the other hand, classification accuracy of 10, 12 and 14 inputs were 5% higher than 2 inputs. It can be concluded that LM training algorithm of ANN was performed better than SCG algorithm with at least 10 inputs. [ABSTRACT FROM AUTHOR]

A passive controlled ankle foot orthosis (PICAFO) used a passive actuator such as Magnetorheological (MR) brake to control the ankle stiffness. The PICAFO used two kinds of sensors, such as Electromyography (EMG) signal and ankle position (two inputs) to determine the amount of stiffness (one output) to be generated by the MR brake. As the overall weight and design of an orthotic device must be optimized, the sensor numbers on PICAFO wanted to be reduced. To do that, a machine learning approach was implemented to simplify the previous stiffness function. In this paper, Non-linear Autoregressive Exogeneous (NARX) neural network were used to generate the simplified function. A total of 2060 data were used to build the network with detail such as 1309 training data, 281 validation data, 281 testing data 1, and 189 testing data 2. Three training algorithms were used such as Levenberg-Marquardt, Bayesian Regularization, and Scaled Conjugate Gradient. The result shows that the function can be simplified into one input (ankle position) – one output (stiffness). Optimized result was shown by the NARX neural network with 15 hidden layers and trained using Bayesian Regularization with delay 2. In this case, the testing data shows R-value of 0.992 and MSE of 19.16. [ABSTRACT FROM AUTHOR]

Featured Application: The control reference parameter suggested in this study is implemented in a Passive-Controlled Ankle Foot Orthosis (PICAFO) equipped with magnetorheological (MR) brake on its ankle joint. Previously, the PICAFO has been shown to successfully deal with foot drop by locking the foot ankle position during the swing phase. By adding functions such as stance assistance, the user can get more benefits when using the PICAFO. This paper aims to present a preliminary study of control reference parameters for stance assistance among different subjects and walking speeds using a passive-controlled ankle foot orthosis. Four young male able-bodied subjects with varying body mass indexes (23.842 ± 4.827) walked in three walking speeds of 1, 3, and 5 km/h. Two control references, average ankle torque (aMa), and ankle angular velocity (aω), which can be implemented using a magnetorheological brake, were measured. Regression analysis was conducted to identify suitable control references in the three different phases of the stance. The results showed that aω has greater correlation (p) with body mass index and walking speed compared to aMa in the whole stance phase (p1(aω) = 0.666 > p1(aMa) = 0.560, p2(aω) = 0.837 > p2(aMa) = 0.277, and p3(aω) = 0.839 > p3(aMa) = 0.369). The estimation standard error (Se) of the aMa was found to be generally higher than of aω (Se1(aMa) = 2.251 > Se1(aω) = 0.786, Se2(aMa) = 1.236 > Se2(aω) = 0.231, Se3(aMa) = 0.696 < Se3(aω) = 0.755). Future studies should perform aω estimation based on body mass index and walking speed, as suggested by the higher correlation and lower standard error as compared to aMa. The number of subjects and walking speed scenarios should also be increased to reduce the standard error of control reference parameters estimation. [ABSTRACT FROM AUTHOR]

This work proposes a new coordinated ambulance routing model suitable for implementation during the COVID-19 pandemic. This model is different from the existing model, where it was conducted uncoordinatedly, so that mismatch between supply and demand may occur. In general, high number of unserved requests and travel distance are unwanted. Therefore, this work proposes a model consisting of three steps: hospital-patient allocation, ambulance-patient dispatching, and ambulance pickup-delivery sequencing. The proposed model consists of two objectives: minimizing the number of unserved patients and minimizing total travel distance. It is developed by using cloud-theory-based simulated annealing. The simulation result shows that the proposed model outperforms the existing uncoordinated model in number of unserved patients, total travel distance, and average travel distance. It creates zero unserved patients if the total number of patients does not surpass the total number of slots in all hospitals. It produces 12 to 19 percent lower total travel distance and 27 to 29 percent lower average travel distance than the uncoordinated model. [ABSTRACT FROM AUTHOR]

Considering persistent years, many researchers continuously seek an optimum way to utilize the idea of magnetorheology (MR) materials to be practically used for everyday life, particularly concerning resistivity sensing application. The rheology and resistivity of a graphite (Gr)-based magnetorheological elastomer (Gr-MRE) were experimentally evaluated in the present research. Magnetorheological elastomer (MRE) samples were prepared by adding Gr as a new additive during MRE fabrication. The effect of additional Gr on the rheological and resistivity properties were investigated and compared with those of typical MREs without a Gr additive. Morphological aspects of Gr-MRE were characterized using field emission scanning electron microscopy (FESEM) and energy dispersive X-ray spectroscopy (EDX). Rheological properties under different magnetic fields were evaluated using a parallel-plate rheometer. Subsequently, the resistivity of all samples was measured under different applied forces and magnetic fields. From the resistivity evaluation, two relationship curves resistance (R) under different applied forces (F) and different magnetic fields (B) were established and plotted by using an empirical model. It was observed from the FESEM images that the presence of Gr fractions arrangement contributes to the conductivity of MRE. It was also observed that, with the addition of Gr, rheological properties such as the field-dependent modulus can be improved, particularly at low strain amplitudes. It is also demonstrated that the addition of Gr in MRE can contribute to the likely use of force detection in tactile sensing devices. [ABSTRACT FROM AUTHOR]

Recently, magnetorheological grease (MRG) has been utilized in magnetorheological (MR) brakes to generate a braking torque based on the current applied. However, the high initial viscosity of MRG has increased the off-state torque that led to the viscous drag of the brake. Therefore, in this study, the off-state viscosity of MRG can be reduced by the introduction of dilution oil as an additive. Three samples consist of pure MRG (MRG 1) and MRG with different types of dilution oil; hydraulic (MRG 2) and kerosene (MRG 3) were prepared by mixing grease and spherical carbonyl iron particles (CIP) using a mechanical stirrer. The rheological properties in the rotational mode were examined using a rheometer and the torque performances in MR brake were evaluated by changing the current of 0 A, 0.4 A, 0.8 A, and 1.2 A with fixed angular speed. The result shows that MRG 3 has the lowest viscosity which is almost 93% reduction while the viscosity of MRG 2 has lowered to 25%. However, the torque performances generated by MRG 3 were highest, 1.44 Nm, when 1.2 A of current was applied and followed by MRG 2 and MRG 1. This phenomenon indicated that the improvement of torque performances was dependent on the viscosity of MRG. By reducing the viscosity of MRG, the restriction on CIP to form chain formation has also decreased and strengthen the torque of MRG brake. Consequently, the utilization of dilution oil in MRG could be considered in MR brake in near future.