Search Results (6)

As cities transition toward intelligent traffic systems, hybrid networks combining AI and traditional intersections raise challenges for efficiency and sustainability. Existing studies primarily focus on global intelligence assumptions, overlooking the practical complexities of hybrid control environments. Moreover, the decision-making processes of AI-based controllers remain opaque, limiting their reliability in dynamic traffic conditions. To address these challenges, this study investigates the following realistic scenario: a Deep Reinforcement Learning (DRL) intersection surrounded by max–pressure-controlled neighbors. A spatiotemporal pressure perception agent is proposed, which (a) uses a novel Holistic Traffic Dynamo State (HTDS) representation that integrates real-time queue, predicted vehicle merging patterns, and approaching traffic flows and (b) innovatively proposes Neighbor–Pressure–Adaptive Reward Weighting (NP-ARW) mechanism to dynamically adjust queue penalties at incoming lanes based on relative pressure differences. Additionally, spatial–temporal pressure features are modeled using 1D convolutional layers (Conv1D) and attention mechanisms. Finally, our Strategy Imitation–Mechanism Attribution framework leverages XGBoost and Decision Trees to systematically analyze traffic condition impacts on phase selection, fundamentally enabling explainable control logic. Experimental results demonstrate the following significant improvements: compared to fixed-time control, our method reduces average travel time by 65.45% and loss time by 85.04%, while simultaneously decreasing average queue lengths and pressure at neighboring intersections by 91.20% and 95.21%, respectively. Full article

Dexterous robotic grasping with multifingered hands remains a critical challenge in non-visual environments, where diverse object geometries and material properties demand adaptive force modulation and tactile-aware manipulation. To address this, we propose the Reinforcement Learning-Based Multimodal Perception (RLMP) framework, which integrates human-like grasping intuition through operator-worn gloves with tactile-guided reinforcement learning. The framework’s key innovation lies in its Tactile-Driven DCNN architecture—a lightweight convolutional network achieving 98.5% object recognition accuracy using spatiotemporal pressure patterns—coupled with an RL policy refinement mechanism that dynamically correlates finger kinematics with real-time tactile feedback. Experimental results demonstrate reliable grasping performance across deformable and rigid objects while maintaining force precision critical for fragile targets. By bridging human teleoperation with autonomous tactile adaptation, RLMP eliminates dependency on visual input and predefined object models, establishing a new paradigm for robotic dexterity in occlusion-rich scenarios. Full article

Migration-induced land degradation is a challenging environmental issue in Sub-Saharan Africa. The need for expansion due to urban development has raised the question of effective sustainable measures. Understanding migration and land degradation links is paramount for sustainable urban development and resource use. This is particularly true in Nigeria, where elevated migration levels frequently result in accelerated land degradation due to urban expansion. Given the need to understand the impact of migration on land degradation in the Savannah Region of Nigeria (SRN), this study introduces a novel approach by integrating remote sensing data (NDVI, NDBI) with local community perceptions (mixed-methods approach) to assess the impact of migration on land degradation in four migration destination communities located in two local government areas (LGAs) (Sabon Gari East and Sabon Gari West of Fagge LGA; Zuba and Tungamaje of Gwagwalada LGA). We conducted focus group discussions and a semi-structured survey with 360 household heads to obtain a comprehensive view of perceptions. Our findings revealed that 41.1% and 29.5% of the respondents agreed and strongly agreed that migration significantly contributes to land degradation. We analysed the spatiotemporal patterns of the Normalised Difference Vegetation Index (NDVI) and the Normalised Difference Built-Up Index (NDBI) acquired from Landsat 8 datasets for 2014 to 2023. While increasing NDBI values were observed in all communities, a slight decrease in NDVI was noted in Sabon Gari East and Tungamaje. Our analyses highlighted activities leading to land degradation such as land pressure due to built-up expansion at Sabon Gari East, Sabon Gari West, and Tungamaje, and deforestation at Zuba. Based on the varying challenges of migration-induced land degradation, we recommend adequate community participation in suggesting targeted interventions and policies to foster various adaptive capacities and sustainable environments within SRN communities and Sub-Saharan Africa. Full article

As the cornerstone of intelligent transportation systems, accurate traffic prediction can reduce the pressure of urban traffic, reduce the cost of residents’ travel time, and provide a reference basis for urban construction planning. Existing traffic prediction methods focus on spatio-temporal dependence modeling, ignoring the influence of weather factors on spatio-temporal characteristics, and the prediction task has complexity and an uneven distribution in different spatio-temporal scenarios and weather changes. In view of this, we propose a weather interaction-aware spatio-temporal attention network (WST-ANet), in which we integrate feature models and dynamic graph modules in the encoder and decoder, and use a spatio-temporal weather interaction perception module for prediction. Firstly, the contextual semantics of the traffic flows are fused using a feature embedding module to improve the adaptability to weather drivers; then, an encoder–decoder is constructed by combining the Dynamic Graph Module and the WSTA Block, to extract spatio-temporal aggregated correlations in the roadway network; finally, the feature information of the encoder was weighted and aggregated using the cross-focusing mechanism, and attention was paid to the hidden state of the encoding. Traffic flow was predicted using the PeMS04 and PeMS08 datasets and compared with multiple typical baseline models. It was learned through extensive experiments that the accuracy evaluation result is the smallest in WST-ANet, which demonstrated the superiority of the proposed model. This can more accurately predict future changes in traffic in different weather conditions, providing decision makers with a basis for optimizing scenarios. Full article

A globally rapid land use/land cover change (LULC) in human-transformed landscapes alters the interface of human-wildlife interactions due to shifting socio-ecological and environmental pressures. Understanding these shifts is crucial for mitigating repeated negative interactions that escalate conflict states between people and wildlife. This study aimed to understand LULC changes over 30 years (1989–2019), with more recent spatio-temporal patterns of high pressure at the human-elephant interface, and potentially underlying environmental and human-driven factors that affect elephant movement patterns. We analyzed a dataset of 923 human-elephant conflict occurrences, mainly crop foraging incidents, in the Enduimet Wildlife Management Area (EWMA) between the years 2016 and 2020 and combined these data with LULC for year 2019 to understand potential drivers of conflict and assess how agricultural land and settlement have increased over time. We further used GPS datasets of elephants collared between 2019 to 2020 to understand elephant movement patterns in changing land use types. Landsat image analysis revealed that 41% of the area had been converted into farmlands and settlements within the last three decades, which creates elephant-intolerant habitats and the potential to increase pressure at the human-elephant interface. Collared elephants using EWMA moved through all land use types and did not avoid settlements, although they moved through these at higher speeds, reflecting perception of risk. Elephants travelled slightly more slowly in farmland, likely reflecting the availability of foraging opportunities. Our analysis shows that human-induced LULC changes and the encroachment into elephant habitats have resulted in spatially and temporally predictable increases in HEC in EWMA, driven by the proximity of farmlands and protected areas (PAs), so that incompatible land uses are the principal drivers of damage to human livelihoods and increased risks to Tanzanian (and Kenyan) natural capital. Communities in Enduimet urgently need support to increase the effective distance between their farming activities and the PAs. Village-level crop protection and small-scale land-use planning around PAs are important first steps to halt an escalating conflict situation but need to be supported with longer-range strategies that separate incompatible land-use types and encourage the cultivation of alternative crops and livelihood diversification. Full article

The study examined the spatiotemporal distribution of drought in the Maasai rangelands of Kenya. The implications of this distribution, in concert with the documented existing and/or projected social and biophysical factors, on critical rangeland resources in Maasai-pastoralism are discussed using an integrated approach. Participatory interviews with the Maasai, retrieval from archives, and acquisition from instrument measurements provided data for the study. Empirical evidence of the current study reveals that drought occurrences in this rangeland have been recurrent, widespread, cyclic, sometimes temporally clustered, and have manifested with varying intensities across spatial, temporal, and, occasionally, social scales; and they have more intensity in lower than higher agroecological areas. An estimated 86% of drought occurrences in this rangeland, over the last three decades alone, were of major drought category. The 2000s, with four major drought events including two extreme droughts, are an important drought period. A strong consensus exists among the Maasai regarding observed drought events. In Maasai-pastoralism, the phenomenon called drought, pastoralist drought, is simultaneously multivariate and multiscalar: its perception comprises the simultaneous manifestation of cross-scale meteorological, socioeconomic, and environmental factors and processes, and their various combinations. The inherent simultaneous multivariate and scalar nature of the pastoralist drought distinguishes it from the conventional drought types, particularly the meteorological drought that predominantly guides drought and resource management in the rangelands of Kenya. In Maasai-pastoralism, the scarcely used (33%) meteorological drought is construed as rainfall delay/failure across spatial and/or temporal scale, and never its reduced amount. Collectively, the current findings reveal that knowledge about drought affects the way the manifestation of this climatic hazard is perceived, communicated, and characterized; hence, ceteris paribus, alongside its spatiotemporal distribution, shapes the nature of the adaptive capacity of and resource management in Maasai-pastoralism. Studies that anticipate enhancing the drought-adaptive capacity of the Maasai should account for cross-scale social and biophysical factors, their processes, and interactions; they must engage the affected inhabitants, and utilize and integrate multiple data sources and approaches. These necessities become more crucial for informing adaptation under the present spatiotemporal distribution of drought as well as in relation to the projected increase in occurrence and intensity of this climatic hazard as the climate continues to change, and as pressures from socioeconomic globalization persistently proliferate into the Maasai’s social and biophysical landscapes. Full article