Narrow Results

The aim of this work is to present a visual-based human action recognition system which is adapted to constrained embedded devices, such as smart phones. Basically, vision-based human action recognition is a combination of feature-tracking, descriptor-extraction and subsequent classification of image representations, with a color-based identification tool to distinguish between multiple human subjects. Simple descriptors sets were evaluated to optimize recognition rate and performance and two dimensional (2D) descriptors were found to be effective. These sets installed on the latest phones can recognize human actions in videos in less than one second with a success rate of over 82%.

The application of artificial intelligence of things (AIoT) is becoming increasingly widespread. With its rapid development, it has raised many complex network security issues. These have brought crushing security threats and risks to all stakeholders. AIoT cloud platform in the smart home plays a vital role in the security management of devices and is the current research hotspot. This paper discusses the security of the AIoT cloud platform, represented by the smart home. The indicator system describing the protection of the smart home cloud platform is proposed, and the security risk evaluation method of the cloud service is formed. Based on the evaluation method, a lightweight security evaluation framework for smart home cloud service is designed to evaluate the typical cloud platform’s security. The evaluation identifies some technical and administrative issues. Finally, to improve the security of the cloud console of AIoT and discover potential security risks in time, a security self-test model of the scalable cloud console is proposed.

In a typical smart home scenario, various household appliances of a residential user such as washing machine and plug-in hybrid electric vehicle (PHEV) are connected via a home area network. Household appliances can be automatically scheduled by a central controller for satisfying the timing and energy constraints. It enables the reduction of monetary cost of electricity consumption and the peak energy load of the home system. On the other hand, the prevailing household appliances usually offer multiple discrete power levels. However, the existing smart home scheduling controllers can only handle the continuous power levels, while makes them unsuitable for prevailing household appliances.

In this paper, a dynamic programming-based scheduling technique is proposed and implemented on FPGA to schedule household appliances with discrete power levels. It features a solution pruning technique, which can largely improve the time complexity. A case study which constitutes ten household appliances is performed. The experimental results demonstrate that our technique can reduce the monetary cost by 39.3% on weekday and 27.2% on weekend, respectively, comparing to the traditional scheduling. For energy consumption balancing, it shows that the peak to average ratio (PAR) of the energy consumption is reduced by 43.6% on weekday and 24.0% on weekend, respectively.

An Internet of Things (IoT) device that can automatically measure water consumption can help prevent excessive water usage or leaks. However, automating too many residences or condominiums with multiple IoT devices can lead to extra energy consumption and more network congestion. We propose controlling the energy consumption of an IoT water consumption management system by dynamically controlling its duty cycle. By analyzing the energy consumption of the developed prototype and its duty cycle variation, we calculated how much energy could be saved by controlling the antenna and the water flow sensor used in the IoT device. While controlling the antenna offered some energy savings, having some way to cut down on the water flow sensor’s consumption can have a dramatic impact on the overall IoT energy consumption or its battery longevity. Our results showed that we could get up to 69% extra energy savings compared to just putting the antenna in sleep mode. There is an observable trade-off in saving so much energy, as we can also see that water reading error rates go up alongside the extra energy savings.

As the world is moving towards Internet of Things, smart home is now rapidly becoming a reality. Home appliances and devices are interconnected, i.e., home area network, via proprietary or standard TCP/IP protocols allowing for better management and monitoring. Nevertheless, as with any form of network, smart home is also prone to security threats and vulnerabilities. This paper aims at presenting the importance of security in smart home environment. Herein, we will discuss (a) the concepts of smart home environment, (b) various smart home communication mechanisms, (c) security challenges and concerns in smart home environment, (d) security threats in smart home environment, (e) current security measures to encounter such security attacks and finally, conclusion and future work.

In the field of product development, there is an increased focus on product complementarity. This study explores how the elements of multi-homing complementarity in the platform ecosystem (complementarity type, degree of platform complementarity, and degree of complementarity with other products) affect perceived product value. The data comprise 248 samples of smart-home multi-homing products, collected from the Amazon EC platform. The results indicate that prioritizing the degree of platform complementarity can increase the number of reviews, while prioritizing the degree of complementarity with other products can reduce that number. Focusing on unique complementarity can have a negative impact on ratings. Regarding the interaction term, the interaction of unique complementarity and the degree of complementarity with other products increases the number of reviews. Thus, unique complementarity should be included in the product-development process to ensure product quality and enhance sales.

Since monitoring one’s ovulation in an accurate manner is significant to the physiological health of women during the reproductive period, it’s better to combine family healthcare and smart home devices to work out feasible ways of ovulation monitoring. A study found that the value of LH(luteinizing hormone) in women’s urine was significantly correlated with ovulation. Therefore, by measuring the peak of LH, the ovulation status of women could be accurately monitored. In this study, a monitoring system was designed to determine ovulation status by detecting LH (luteinizing hormone) value in female urine. This monitoring system was integrated with the household toilet to realize the real-time and accurate monitoring of ovulation status of female individual. In the aspect of hardware, the device appearance, the detection device structure and the micro-control electronic device were designed by applying the Internet of things technology, to help women obtain ovulation information in a convenient and stable manner. Whereas in the aspect of software, the author analyzed the monitoring data based on computer algorithm, developed mobile application, and established the individual longitudinal database with multiple detection values as background data, so as to predict ovulation activity and evaluate ovulation patterns. It would help women track their ovulation more accurately, prevent and indicate gynecological reproductive diseases, and can serve as a supplementary basis for medical diagnosis and treatment results.

This paper present a system for temperature monitoring and command based on GSM networks which can be used in smart home. The system made use of GSM networks, the temperature controller, transducers and LabVIEW to achieve remote monitoring. The user can set the temperature by sending a message and be informed by receiving an over-temperature alarm message. The result shows that this system not only can be controlled by GSM and set a given temperature, but also can automatically send the alert message or make an emergency phone call to a given telephone number

The application of non-specific voice identification technology has been studied in the smart home system. Combining with the related technologies and features of Internet of things, the voice control system of smart home has been designed. The mobile terminal control system (supporting Android and IOS systems) and embedded portable voice controller are made up of the system; the voice cloud technology is adopted in mobile terminal control software, the embedded portable voice terminal is designing by the systematic framework of LD3320, STM32 and CC2530, combining with IEEE 802.15.4 protocol. The practice verifies that the voice control system implements diversified and comprehensive voice control for controllable devices in smart home.

Wireless network based is a trend in the development of smart home control, but the equipment is not compatible or has a unified communication protocol restrict its development. According to these two problems, this paper proposes a wireless communication protocol which is suitable for smart home networked control system. The protocol architecture, interaction and frame format are given. The CPN model of the control system is established, and the activity of the system is verified, and the system transmission delay characteristic is simulated. Verification and simulation results show that the system has no deadlock, and the transmission delay meet the system requirements.