Predicting large-scale transportation network traffic has become an important and challenging topic in recent decades. Inspired by the domain knowledge of motion prediction, in which the future motion of an object can be predicted based on previous scenes, we propose a network grid representation method that can retain the fine-scale structure of a transportation network. Network-wide traffic speeds are converted into a series of static images and input into a novel deep architecture, namely, spatiotemporal recurrent convolutional networks (SRCNs), for traffic forecasting. The proposed SRCNs inherit the advantages of deep convolutional neural networks (DCNNs) and long short-term memory (LSTM) neural networks. The spatial dependencies of network-wide traffic can be captured by DCNNs, and the temporal dynamics can be learned by LSTMs. An experiment on a Beijing transportation network with 278 links demonstrates that SRCNs outperform other deep learning-based algorithms in both short-term and long-term traffic prediction.

No abstract

Studies on the azobenzene derivative based phase transitions mostly rely on photoisomerization, which require a long time to spontaneously revert back. Here we show a photothermal-driven solid-to-liquid transition and fast reversion of azobenzene derivatives. Owing to the aggregation of suitably substituted azobenzenes, solid-to-liquid transitions can be induced by photothermal effects under irradiation with green light. The liquid-state azobenzene derivatives spontaneously solidify again within 2 min due to heat release in a purely physical fashion. One thus obtains a perfectly reversible adhesion with a strength as high as that of commercial materials. Our work affords a novel concept to construct reversible adhesives via phase transitions of organic compounds induced by light.

As ice plays a critical role in various aspects of life, from food preservation to ice sports and ecosystem, it is desirable to protect ice from melting, especially under sunlight. The fundamental reason for ice melt under sunlight is related to the imbalanced energy flows of the incoming sunlight and outgoing thermal radiation. Therefore, radiative cooling, which can balance the energy flows without energy consumption, offers a sustainable approach for ice protection. Here, we demonstrate that a hierarchically designed radiative cooling film based on abundant and eco-friendly cellulose acetate molecules versatilely provides effective and passive protection to various forms/scales of ice under sunlight. This work provides inspiration for developing an effective, scalable, and sustainable route for preserving ice and other critical elements of ecosystems.