Search

Abstract Fires in petrochemical complexes are inaccessible because of the intense radiant heat from flames. Therefore, water cannon robots require radiant heat countermeasures to perform firefighting safely. Conventional radiant heat countermeasures employ a self-spraying method, wherein the water cannon robot requires a water tank of capacity 1.5 m3 (= 1500 L) to function for 7–8 min in an environment with a 20 kW/m2 radiation heat. However, the water cannon robot has size limitations because it is transported on one transport vehicle (10 t truck) to the site, and only a tank of capacity ~ 0.02 m3 can be installed on the robot. To overcome these drawbacks, this study proposes a method that utilizes a mountable radiant heat-resistant exterior cover that works with a small amount of water. The cover is made of radiant heat-shielding fireproof clothing with an aluminum coating that reflects 90% of the radiant heat on the surface and a mist nozzle that sprays water on the back surface. The remaining 10% is removed by the heat of vaporization of water sprayed on the back of the clothing and natural convection. The amount of water required for cooling was reduced to 1/80th of that compared to self-spraying because of the use of the developed cover. The proposed method of radiation reflection via vaporization and natural convection can be employed to protect firefighting robots.

Abstract Citizen science programs using organism photographs have become popular, but there are two problems related to photographs. One problem is the low quality of photographs. It is laborious to identify species in photographs taken outdoors because they are out of focus, partially invisible, or under different lighting conditions. The other is difficulty for non-experts to identify species. Organisms usually have interspecific similarity and intraspecific variation, which hinder species identification by non-experts. Deep learning solves these problems and increases the availability of organism photographs. We trained a deep convolutional neural network, Xception, to identify bee species using various quality of bee photographs that were taken by citizens. These bees belonged to two honey bee species and 10 bumble bee species with interspecific similarity and intraspecific variation. We investigated the accuracy of species identification by biologists and deep learning. The accuracy of species identification by Xception (83.4%) was much higher than that of biologists (53.7%). When we grouped bee photographs by different colors resulting from intraspecific variation in addition to species, the accuracy of species identification by Xception increased to 84.7%. The collaboration with deep learning and experts will increase the reliability of species identification and their use for scientific researches.