Search for: All records

Evaporative misters have long been used in urban spaces for heat mitigation, yet their thermal stress impacts and optimal operating conditions have not been fully explored. To fill this gap, we develop a misting model and embed it into an urban canopy model for the first time. Our tests confirm that misters can considerably reduce maximum urban canyon air temperature (up to 17.5 °C) and human skin temperature (up to 0.48 °C) in a hot and dry city (Phoenix, AZ). They continue to effectively reduce thermal stress, albeit with half of the cooling benefits, in a hot and humid city (Houston, TX). These thermal stress impacts are contingent upon wind speeds: the optimal wind speeds generally fall within an intermediate range—from light air (with low mist flow rates) to a moderate breeze (with higher mist flow rates). We then incorporate misting into a broader comparison of blue cooling strategies, including irrigation (on vegetation) and sprinkling (on pavements). With abundant water resources, sprinkling on asphalt and misting are the most effective cooling solutions, particularly suitable for middays and late afternoons, respectively. To balance cooling benefits with limited water resources, we propose a thermostatic control scheme that can save at least 10.5 m3/day of water compared to continuous misting for a 100-m stretch of street, equivalent to the water demand of about 20 Phoenix residents. Notably, misting and sprinkling generate rapid cooling in under 10 min with sufficient flow rates, demonstrating their potential as fast activation measures during extreme heat emergencies. 

Visual contents, including images and videos, are dominant on the Internet today. The conventional search engine is mainly designed for textual documents, which must be extended to process and manage increasingly high volumes of visual data objects. In this paper, we present Mixer, an effective system to identify and analyze visual contents and to extract their features for data retrievals, aiming at addressing two critical issues: (1) efficiently and timely understanding visual contents, (2) retrieving them at high precision and recall rates without impairing the performance. In Mixer, the visual objects are categorized into different classes, each of which has representative visual features. Subsystems for model production and model execution are developed. Two retrieval layers are designed and implemented for images and videos, respectively. In this way, we are able to perform aggregation retrievals of the two types in efficient ways. The experiments with Baidu's production workloads and systems show that Mixer halves the model production time and raises the feature production throughput by 9.14x. Mixer also achieves the precision and recall of video retrievals at 95% and 97%, respectively. Mixer has been in its daily operations, which makes the search engine highly scalable for visual contents at a low cost. Having observed productivity improvement of upper-level applications in the search engine, we believe our system framework would generally benefit other data processing applications.