No abstract

Road crashes are a growing concern of governments and is rising to become one of the leading preventable causes of death, especially in developing countries. The ubiquitous presence of smartphones provides a new platform on which to implement sensor networks and driver assistance systems, as well as other ITS applications. In this paper, existing approaches of using smartphones for ITS applications are analyzed and compared. Particular focus is placed on vehicle-based monitoring systems, such as driving behavior and style recognition, accident detection and road condition monitoring systems. Further opportunities for use of smartphones in ITS systems are highlighted, and remaining challenges in this emerging field of research are identified.

Faced with the threat of "Day Zero", when it was feared that Cape Town's taps could run dry, consumers reduced household water usage from 540 to 280 litres per household per day over the 36 months between January 2015 and January 2018. This paper describes the events that prompted this reduction. We look at how changes in water use were affected by official announcements and by public engagement with this news via the social media activity and internet searches. We analysed the water usage of a subset of middle to high income households where smart hot and cold water meters were installed. For hot water usage patterns we compared meter readings with that in another area unaffected by the drought. We further map our cold water smart meter readings against that of the City of Cape Town's municipal data for domestic freestanding households-a sample of more than 400,000 households. We found that the introduction of Level 5 restrictions had a perverse effect on consumption, possibly due to confusing messages. The most dramatic change in behaviour appears to have been instigated by a media storm and consequent user panic after the release of the City's Critical Water Shortages Disaster Plan in October 2017. However, contradictory communication from national and provincial government eroded some of this gain. The paper concludes with recommendations for demand management in a similar future scenario.

Recent advances in smart grid technology enable new approaches to address the problem of load control for domestic water heating. Since water heaters store energy, they are well-suited to load management. However, existing approaches have focused on the electrical supply side, ignoring the obvious link between the user and the grid: individual hot water consumption patterns. This paper proposes a load spreading approach in which water heaters compete for access to the heating medium. The proposed smart grid solution takes grid load limits, real-time temperature measurements, water usage patterns, individual user comfort, and heater meta-data into consideration. The scheduler only turns on the heaters with the highest level of need, but limits the number of on heaters to ensure that the grid load stays below a set limit for a set time. The method is evaluated by simulation against various heater set temperature levels, and for various load limits, and compared with ripple control and actual consumption measured in a field trial of 34 water heaters. The proposed algorithm reduces the load from 62kW to 20, 30, 40, and 50kW (vs. 106kW for full ripple control). The resulting number of unwanted cold events is fewer than for ripple control, and only slightly more than no control, while reducing the total energy by 14% from a user-optimised natural experiment.

Electric water heaters (EWHs) remain one of the main contributors to energy consumption in countries where they are used. EWH models serve as a step towards achieving optimised control, and can also be used to inform users of expected savings due to changes, if the model is energy-based. Various models have been proposed, but none of them include more than half of the six key features that the model presented in this paper supports: horizontal orientation; schedule control; low computational complexity; validation of the model; multinodal stratification; and multinodal standing losses. The presented model is validated against six datasets: four comprising 900 hours with multiple water usage events; and two with only standing losses. The results show that the model estimates energy consumption over ten days including usage with an error of less than 2% and 5% for schedule control and thermostat control respectively. The simulation model is simple enough to execute ten days of simulation in less than 100 milliseconds on a standard desktop machine, 150 times faster than a prominent model from literature, making it also suitable for large scale simulations or for use on mobile devices.