Fig 4 - uploaded by Marcin Badurowicz
Content may be subject to copyright.
Source publication
The paper presents the cloud computing system designed for monitoring the state of roads by processing data packages covering such data as the car’s acceleration and position acquired by mobile devices (smartphones and tablets) mounted in cars and implemented on IBM BlueMix platform. Such data are being directly sent to the cloud system, where they...
Context in source publication
Context 1
... previous research it was calculated that best results are achieved with a threshold of 4.3 times the standard deviation (4.3σ). In Figure 4 below there is a standard deviation times threshold presented over the calculated Z s data to show that three data points (at 6.2, 6.5 and 6.8 seconds, respectively) are above the dened threshold and will be recognised as road artefacts. This causes two issues. ...
Citations
... The device was accompanied by a second test device, a Nokia Lumia 820 smartphone equipped with the previous implementation of the author's data acquisition software (Badurowicz et al., 2016), which will be used to cross-validate the results. ...
In the paper, the authors are presenting the analysis of implementation of IoT system of road quality analysis. The proposed system has been prepared with edge, on-device processing in mind, allowing for reduction of amount of data being sent to cloud computing aggregation subsystem, sending only 2.5% of the original data. Several algorithms for road quality analysis has been implemented on a real device and tested in a real-world conditions. The system has been compared to the state-of-the-art offline processing approach and shown very similar results.
... In [1] it was proposed a modification to the previous methods by usage of Grubbs test. The authors earlier already implemented a modified version of the Z-THRESH method, called MOD-Z-THRESH [8], where threshold value was calculated relatively to the overall road surface quality in a set of tumbling windows, instead of using strictly defined threshold. Other threshold-based methods include mostly also just Z-axis acceleration [9], but it must be noted that some more complex features extracted from accelerometer signals, for example time domain features, such as mean, median etc. [10], could be used in the described problem. ...
... Another techniques accuracy is ranging from 91.43% in the case of the rough road [18], to 94% when Artificial Neural Networks and similar techniques are used [19]. The authors' own method MOD-Z-THRESH [8] can provide an average accuracy of about 93.2%. ...
... Because of difference between cars and their parameters, every data can be understood as averaged fuzzy information. The data acquisition procedure was based on the concept already implemented in [8], where smartphone, Lumia 820, was mounted in the car in a stable position, which was the only requirement, and in case of the experiments presented in this paper it was in the central console. ...
... Threshold-based methods are the most straightforward approaches for PD detection by processing mainly the vertical acceleration (Z-acc) or in combination with other direction acceleration (x and y) and gyroscopes. A researcher [62] has proposed four indices in which the Z-THRESH was further modified [63] to build a cloud computing system: Z-THRESH (from vertical vibration), Z-DIFF (from the difference of consecutive Z-acc above threshold), STDEV(Z) (as the standard deviation of Zacc above threshold in a window), and G-ZERO (whether the sensor senses a 0-G vibration). Similar STDEV(Z) can also be found in [64] and to develop a bump index in [65]. ...
With the development of smart technologies, Internet of Things and inexpensive onboard sensors, many response-based methods to evaluate road surface conditions have emerged in the recent decade. Various techniques and systems have been developed to measure road profiles and detect road anomalies for multiple purposes such as expedient maintenance of pavements and adaptive control of vehicle dynamics to improve ride comfort and ride handling. A holistic review of studies into modern response-based techniques for road pavement applications is found to be lacking. Herein, the focus of this article is threefold: to provide an overview of the state-of-the-art response-based methods, to highlight key differences between methods and thereby to propose key focus areas for future research.
The results show that machine-learning techniques/data-driven methods have been used intensively with promising results but the disadvantages on data dependence have limited its application in some instances as compared to analytical/data processing methods. Recent algorithms to reconstruct/estimate road profiles are based mainly on passive suspension and quarter-vehicle-model, utilise fewer key parameters, being independent on speed variation and less computation for real-time/online applications. On the other hand, algorithms for pothole detection and road roughness index estimation are increasingly focusing on GPS accuracy, data aggregation and crowdsourcing platform for large-scale application. However, a novel and comprehensive system that is comparable to existing International Roughness Index and conventional Pavement Management System is still lacking.
... A motion recognition and collision detection using the acceleration and magnetic fields sensors embedded in smartphones is presented in [20]. The sensors are also used for collision detection [2]. The sensors, like accelerometers, produce inaccurate noisy data. ...
The aim of the paper is to compare two methods of motion data acquisition using: (1) a mobile device, and (2) an optical reference system. The paper presents a mobile application developed by the authors for the Android platform which was used for motion registration process. The application reads the data from the embedded accelerometer and magnetometer sensors in three dimensions (along X, Y and Z axis). The application performance was evaluated with the help of a passive motion capture system, which was used as a reference. The presented analysis indicates how precise the mobile registration method is in relation to the reference system. Distance and speed are the parameters that were taken into the consideration. Motion was registered by a mobile device attached to the participant’s arm. Retroreflective markers, required by the reference system were attached to the phone and mounting bands. The participant performed the following activities: walking and running. The results obtained using mobile devices were not precise and have been found to strongly depend on the mobile device used. They may be useful for gathering coarse motion statistics.
... Accelerometer Normalisation Factor is being calculated when the test car is driven over the known road segment with the same speed as the baseline car. The road segment is carefully selected as one of the A-class road segments, for example newly built highways [11]. First, every sample of dataset (X) to be normalised (x) are scaled into values from the set between minimum and maximum values for the baseline (Y), resulting in normalised x (x n ). ...
... To confront validity of the constructed virtual road against its real world counterpart, the authors used the same procedure for assessing the virtual road quality as for the real roads -the Road Relative Unevenness Index (RRUI) was used [11]. ...
Road quality assessment using crowdsourced data gathered by smartphone users, based on acceleration data, is an interesting subject on using modern technology for improvements of the infrastructure. The algorithms – for both road quality assessment and detection of different elements on the road – need to be tested, especially in the field. To facilitate building sets of different data and sharing them in a standardised way, the authors propose extraction of known road fragments with known types of surface degradation and construction of virtual streams of data, thus “virtual roads”. The procedure for data extraction and building a database of segments, combining them into virtual road, as well as testing real-world algorithm using the constructed virtual road are presented in the paper.
... Since the accepting gesture should be intuitive and fast, the authors conducted basic experiments to evaluate the possibility of using it. Using the developed software, acquiring roll, pitch and yaw values every 0.015 of a second, by performing similar data acquisition and postprocessing procedure as already established in another research [14], the accepting gesture was experimentally validated, using the software prototype. In the fig. 2 the prototype software (version 1) during the acquisition phase of the inclinometer values is presented. ...
In the paper authors are introducing the concept of usage of physical orientation of a mobile device, calculated using built-in environmental sensors like accelerometer, gyroscope and magnetometer for detection of tilting gesture. This gesture is used as an acceptance factor for the two next probable word solutions suggested to the user during text input. By performing the device tilt, the first or second word is being automatically put into the desired text field and new prediction is performed. The text predictions are calculated and stored directly on the device to maintain privacy protection. The founding concept of the software is being presented, as well as initial considerations and further plans. This solution is recommended especially to smartphone manufacturers like Microsoft, Samsung and Apple to deploy in their latest models.
The development of numerical methods and information technologies is occurring at a rapid pace. Hence, the main research question addressed by this work is whether ethics is required for artificial intelligence (AI), and thus (1) ‘Are the existing documents regulating the ethics of AI sufficient?’ and (2) ‘Is there a need to develop new regulations? If so, in what areas?’ The presented investigations involved the use of literature review method, case study, analysis, and synthesis. Based on the analysis of scientific documents and articles, as well as examples related to the use of artificial intelligence, the main and detailed research questions have been answered in the affirmative. Nonetheless, there is a pressing need to develop documentation regulating: military AI use, AI applications in social networks, robot ethics, AI in the automotive industry, or ecology. The presented case studies indicate other unregulated areas: AI applications in medicine or legal provisions on specific issues. The lack of an AI code of ethics may hinder the development of new applications for intelligent devices in the future.
