Hasil untuk "Regional planning"

S. Fragomeni, A. Sciallis, J. Jeruss

Lei Shi, Victor Aregbede, Andreas Persson et al.

Procedure planning requires a model to predict a sequence of actions that transform a start visual observation into a goal in instructional videos. While most existing methods rely primarily on visual observations as input, they often struggle with the inherent ambiguity where different actions can appear visually similar. In this work, we argue that language descriptions offer a more distinctive representation in the latent space for procedure planning. We introduce Language-Aware Planning (LAP), a novel method that leverages the expressiveness of language to bridge visual observation and planning. LAP uses a finetuned Vision Language Model (VLM) to translate visual observations into text descriptions and to predict actions and extract text embeddings. These text embeddings are more distinctive than visual embeddings and are used in a diffusion model for planning action sequences. We evaluate LAP on three procedure planning benchmarks: CrossTask, Coin, and NIV. LAP achieves new state-of-the-art performance across multiple metrics and time horizons by large margin, demonstrating the significant advantage of language-aware planning.

Yarin Benyamin, Argaman Mordoch, Shahaf S. Shperberg et al.

Large Language Models (LLMs) are increasingly being used as autonomous agents capable of performing complicated tasks. However, they lack the ability to perform reliable long-horizon planning on their own. This paper bridges this gap by introducing the Planning Copilot, a chatbot that integrates multiple planning tools and allows users to invoke them through instructions in natural language. The Planning Copilot leverages the Model Context Protocol (MCP), a recently developed standard for connecting LLMs with external tools and systems. This approach allows using any LLM that supports MCP without domain-specific fine-tuning. Our Planning Copilot supports common planning tasks such as checking the syntax of planning problems, selecting an appropriate planner, calling it, validating the plan it generates, and simulating their execution. We empirically evaluate the ability of our Planning Copilot to perform these tasks using three open-source LLMs. The results show that the Planning Copilot highly outperforms using the same LLMs without the planning tools. We also conducted a limited qualitative comparison of our tool against Chat GPT-5, a very recent commercial LLM. Our results shows that our Planning Copilot significantly outperforms GPT-5 despite relying on a much smaller LLM. This suggests dedicated planning tools may be an effective way to enable LLMs to perform planning tasks.

Alberto Pozanco, Marianela Morales, Daniel Borrajo et al.

Identifying the specific actions that achieve goals when solving a planning task might be beneficial for various planning applications. Traditionally, this identification occurs post-search, as some actions may temporarily achieve goals that are later undone and re-achieved by other actions. In this paper, we propose a compilation that extends the original planning task with commit actions that enforce the persistence of specific goals once achieved, allowing planners to identify permanent goal achievement during planning. Experimental results indicate that solving the reformulated tasks does not incur on any additional overhead both when performing optimal and suboptimal planning, while providing useful information for some downstream tasks.

David Speck, Markus Hecher, Daniel Gnad et al.

Classical planning asks for a sequence of operators reaching a given goal. While the most common case is to compute a plan, many scenarios require more than that. However, quantitative reasoning on the plan space remains mostly unexplored. A fundamental problem is to count plans, which relates to the conditional probability on the plan space. Indeed, qualitative and quantitative approaches are well-established in various other areas of automated reasoning. We present the first study to quantitative and qualitative reasoning on the plan space. In particular, we focus on polynomially bounded plans. On the theoretical side, we study its complexity, which gives rise to rich reasoning modes. Since counting is hard in general, we introduce the easier notion of facets, which enables understanding the significance of operators. On the practical side, we implement quantitative reasoning for planning. Thereby, we transform a planning task into a propositional formula and use knowledge compilation to count different plans. This framework scales well to large plan spaces, while enabling rich reasoning capabilities such as learning pruning functions and explainable planning.

Mikhail Soutchanski, Yongmei Liu

In classical planning and conformant planning, it is assumed that there are finitely many named objects given in advance, and only they can participate in actions and in fluents. This is the Domain Closure Assumption (DCA). However, there are practical planning problems where the set of objects changes dynamically as actions are performed; e.g., new objects can be created, old objects can be destroyed. We formulate the planning problem in first-order logic, assume an initial theory is a finite consistent set of fluent literals, discuss when this guarantees that in every situation there are only finitely many possible actions, impose a finite integer bound on the length of the plan, and propose to organize search over sequences of actions that are grounded at planning time. We show the soundness and completeness of our approach. It can be used to solve the bounded planning problems without DCA that belong to the intersection of sequential generalized planning (without sensing actions) and conformant planning, restricted to the case without the disjunction over fluent literals. We discuss a proof-of-the-concept implementation of our planner.

Patrick R. Brown, A. Botterud

Summary Preventing global warming in excess of 1.5°C–2°C requires a transition to zero-carbon electricity systems by midcentury along with the widespread electrification of other sectors. Current state-level renewable portfolio standards and regional transmission arrangements do not capture the benefits of inter-regional transmission or coordination of planning and dispatch for renewable-energy integration. Here, using a co-optimized capacity-planning and dispatch model over 7 years of hourly operation, we show that inter-state coordination and transmission expansion reduce the system cost of electricity in a 100%-renewable US power system by 46% compared with a state-by-state approach, from 135 $/MWh to 73 $/MWh. Sensitivity analyses show that reductions in the cost of photovoltaics, wind, and lithium-ion batteries lead to the lowest electricity costs for systems in which transmission expansion is allowed, while cost reductions for nuclear power or long-duration energy storage lead to greater electricity cost reductions for isolated systems.

Guangsheng Pan, W. Gu, Haifeng Qiu et al.

Abstract Hydrogen is regarded as secondary energy that is perfectly complementary to electricity owing to its friendly storage characteristics and can play a vital role in the future low-carbon society. Toward that end, we propose a regional electricity-hydrogen integrated energy system that can achieve high penetration of renewable energy using electricity and hydrogen as energy carriers. A bi-level mixed-integer planning model is proposed to highlight the role of hydrogen in renewable energy penetration and seasonal complementarity. The upper-level model aims at improving the system economy and optimizes the equipment configuration to meet the regional energy demands; the lower-level model minimizes the levelized cost of hydrogen to promote the development of hydrogen. Both the two levels cover binary variables to characterize the interactive states and ON/OFF states, which makes the bi-level model cannot be directly translated into an equivalent mathematical program with equilibrium constraints problem. Then, a reformulation and decomposition algorithm is applied to handle this complex problem with limited iterations. Case studies show that the proposed model can achieve the dual goals of optimizing the equipment configuration and reducing the supply price of hydrogen by rationally using resources such as wind, solar, and geothermal energy in the planning stage.

Vahid Isazade, Abdul Baser Qasimi, Abdulla Al Kafy et al.

Flood events are the most sophisticated and damaging natural hazard compared to other natural catastrophes. Every year, this hazard causes human-financial losses and damage to croplands in different locations worldwide. This research employs a combination of artificial neural networks and geographic information systems (GIS) to simulate flood-vulnerable locations in the Samangan Province of Afghanistan. First, flood-influencing factors, such as soil, slope layer, elevation, flow direction, and land use/cover, were evaluated as influential factors in simulating flood-prone areas. These factors were imported into GIS software. The Fishnet command was used to partition the information layers. Furthermore, each layer was converted into points, and this data was fed into the perceptron neural network along with the educational data obtained from Google Earth. In the perceptron neural network, the input layers have five neurons and 16 nodes, and the outputs showed that elevation had the lowest possible weight (R2 = 0.713) and flow direction had the highest weight (R2 = 0.913). This study demonstrated that combining GIS and artificial neural networks results in acceptable performance for simulating and modeling flood susceptible areas in different geographical locations and significantly helps prevent or reduce flood hazards.

Xiaomin Wang, Zhiwei Ding

With the strengthening of the cross-regional flows of the economy, information, innovation, and population, this paper constructs a network model of multi-flow integration and analyzes the spatial pattern and influencing factors of urban networks in Chengdu-Chongqing Urban Agglomeration using social network analysis and spatial analysis technology. The main conclusions are as follows. (1) The density and efficiency are in the transition stage from the primary level to the medium level in the comprehensive network. (2) The overall pattern keeps a polyhedral pyramid structure with Chengdu ↔ Chongqing as the core axis, and the grade of each axis has been significantly raised. (3) Four groups are formed using the social network method and show a geographic proximity effect. In addition, the connections within each group are relatively close, but the connections between the groups are significantly different. (4) Location conditions, economic development level, enterprise development level, scientific research investment, scientific and technological development level, and government support have a greater impact on the formation of the comprehensive network of Chengdu-Chongqing urban agglomeration. Information application level and transportation accessibility show a small impact and human capital level has not yet produced a significant impact.

E. Goetz, Rashad Williams, A. Damiano

Abstract Problem, research strategy, and findings: The ability of planning to address America’s urban problems of inequality, crime, housing, education, and segregation is hampered by a relative neglect of Whiteness and its role in shaping urban outcomes. We offer a justification for centering Whiteness within urban planning scholarship and practice that would examine its role shaping and perpetuating regional and racial injustices in the American city. The focus of planners, scholars, and public discourse on the “dysfunctions” of communities of color, notably poverty, high levels of segregation, and isolation, diverts attention from the structural systems that produce and reproduce the advantages of affluent and White neighborhoods. Planners and planning scholars frequently invoke a “legacy of injustice” with regard to concentrated poverty and disadvantage but not in regard to neighborhoods of White affluence. One is segregated and problematized and the other is idealized. Takeaway for practice: Planners and planning scholars need to understand the role of Whiteness, in particular White affluence, to assess the potential impacts of planning interventions. Doing so will inform a wider range of planning approaches to problems of racial and spatial equity.

V. Nadin, D. Stead, M. Dąbrowski et al.

ABSTRACT Whether spatial planning systems are equipped to cope with contemporary regional and urban challenges is strongly dependent on their capacity to promote integration between policy sectors, to respond adaptively to changing societal and political conditions, and to involve and engage citizens in decision-making processes. This paper examines and compares how these capacities have evolved in European countries since the start of the 21st century. The findings indicate that many countries have made reforms to spatial planning with significant implications for their capacity to promote integrated, adaptive and collective planning decisions.

Yaohua Cheng, Ning Zhang, Zongxiang Lu et al.

Coordinating multiple energy systems (MES) enables the synergies of different energy sectors to be exploited. The concept of low-carbon development changes planning approaches for both the district level and the multi-regional level of MES. This paper proposes a bi-level expansion planning model of MES that considers the emission constraints under a decentralized approach. The upper-level model investigates the optimal planning scheme for integrated power and natural gas networks in the multi-regional MES. The lower-level model examines the optimal energy supply configuration of multiple energy carriers in the district MES based on the energy hub modeling approach. The carbon emission flow model is used to allocate the overall carbon emission cap among district MES from the consumption-based perspective and to coordinate the planning of the district level and the multi-regional level. An illustrative example based on a 6-node MES verifies the effectiveness of the proposed model. Finally, a realistic case study based on an MES in northern China is implemented to show the effects of carbon emission constraints on the planning of real-world energy systems.

W. Verheye

Jinsong Li, Shaochen Wang, Ziyang Chen et al.

Learning-based path planning is becoming a promising robot navigation methodology due to its adaptability to various environments. However, the expensive computing and storage associated with networks impose significant challenges for their deployment on low-cost robots. Motivated by this practical challenge, we develop a lightweight neural path planning architecture with a dual input network and a hybrid sampler for resource-constrained robotic systems. Our architecture is designed with efficient task feature extraction and fusion modules to translate the given planning instance into a guidance map. The hybrid sampler is then applied to restrict the planning within the prospective regions indicated by the guide map. To enable the network training, we further construct a publicly available dataset with various successful planning instances. Numerical simulations and physical experiments demonstrate that, compared with baseline approaches, our approach has nearly an order of magnitude fewer model size and five times lower computational while achieving promising performance. Besides, our approach can also accelerate the planning convergence process with fewer planning iterations compared to sample-based methods.

Ofori Prince Danso, Bismark Asante-Badu, Zezhou Zhang et al.

Selenium (Se) is an essential trace element for humans and animals. Its necessity for plants is still under examination. Due to the contradictory nature of Se and its significance, it has received much interest in recent years. Se deficiency can be harmful to humans, yet almost a billion people are deficient. Its deficiency has been associated with cancers, impairment of organs, and a number of other ailments. The biofortification of plants and livestock is a guaranteed practice to increase human selenium consumption. Strategies such as foliar spraying, the direct application of Se in plants and Se feed, and injections in livestock have been employed. Se biofortification has been shown to have additional beneficial effects in plants and livestock. In plants, it has been reported to mitigate different types of stress and increase yield. In animal biofortification, Se has been shown to reduce the detrimental effects of ailments and promote healthy growth. Se biofortification, nevertheless, confronts a number of difficulties. For instance, the bulk of biofortified products must be prepared before consumption, lowering the Se concentration. The objective of this review is to convey the current understanding of the Se biofortification of plants and animals, as well as its difficulties, taking into account both the detrimental consequences of Se deficiency and benefits of Se biofortification.

Anna Maria Tri Anggraini, Maya Indrasti Notoprayitno

The existence of a consumer protection law and the ratification of the CRPD by the Indonesian government is expected to guarantee the safety and comfort of consumers, including persons with disabilities. Therefore, the problem of comparative regulation and institutional, as well as the supervision of the implementation of public services for persons with disabilities, is raised in Indonesia and Australia. Australia was chosen as a comparison because this country already has a comprehensive protection system for persons with disabilities and is fully committed to providing public service facilities. This research is a prescriptive normative research using secondary data consisting of primary legal materials and secondary legal materials. This study concludes that similar to Indonesia, the formation of regulations and policies in Australia in the public service sector for persons with disabilities has reached a technical level and is carried out in a coordinated manner between the center and the regions. The basic difference is that the institutional system that handles the planning, implementation, and supervision of public services for persons with disabilities in Indonesia is separated into various ministries and/or agencies so that it requires strengthening synergies at the central and regional levels so that the implementation of public services is guaranteed optimally.

E. Alexander

The futility of defining planning suggests that there is no planning as a recognizable practice. Sociology of knowledge definitions imply three kinds of planning practices: (1) Generic “planning”—what people do when they are planning; (2) Knowledge-centered “something” (e.g., spatial) planning; and (3) Real planning practiced in specific contexts, from metro-regional planning for Jakarta to transportation planning for the Trans-Europe Network, and enacted in general contexts, for example, informal- or Southern planning. Planning theories are linked to different practices: generic “planning” theories and “something” (e.g., regional, community, environmental, or Southern) planning theories. Selected topics illustrate the “planning” theory discourse and spatial planning theories are briefly reviewed. Three generations of planning practice studies are reviewed: the first, a-theoretical; the second, the “practice movement,” who studied practice for their own theorizing; and the third, informed by practice theories. Five books about planning show how their planning theorist authors understand planning practice. While recognizing planning as diverse practices, they hardly apply “planning” theory to planning practices. “Planning” theories are divorced from enacted planning practices, “something” (e.g., spatial) planning theories include constructive adaptations of “planning” theories and paradigms, but knowledge about real planning practices is limited. Implications from these conclusions are drawn for planning theory, education, and practices.

P. Angelstam, Michael Manton, Martin Green et al.

Abstract Loss of forest naturalness challenges the maintenance of green infrastructure (GI) for biodiversity conservation and delivery of diverse ecosystem services. Using the Convention on Biological Diversity’s Aichi target #11 with its quantitative and qualitative criteria as a normative model, we aim at supporting landscape planning through a pioneering assessment of the extent to which existing amounts and spatial distributions of High Conservation Value Forests (HCVFs) meet these criteria. Highly forested and committed to both intensive wood production and evidence-based conservation targets of 17–20% protected areas, Sweden was chosen as a case study. Specifically, we estimated the amount, regional representation, and functional connectivity of HCVF patches using virtual bird species, validated the results using field surveys of focal bird species, and assessed conservation target fulfilment. Finally, we linked these results to the regional distribution of forest land ownership categories, and stress that these provide different opportunities for landscape planning. Even if 31% of forest land in Sweden is officially protected, voluntarily set-aside, or not used for wood production now and in the future, we show that applying the representation and connectivity criteria of Aichi target #11 reduces this figure to an effective GI of 12%. When disaggregating the five ecoregions the effective GI was 54% for the sub-alpine forest ecoregion, which hosts EU’s last intact forest landscapes, but only 3–8% in the other four ecoregions where wood production is predominant. This results in an increasing need for forest habitat and landscape restoration from north to south. The large regional variation in the opportunity for landscape planning stresses the need for a portfolio of different approaches. We stress the need to secure funding mechanisms for compensating land owners’ investments in GI, and to adapt both the approaches and spatial extents of landscape planning units to land ownership structure.

Beomjoon Kim, Luke Shimanuki, Leslie Pack Kaelbling et al.

We present a framework for learning to guide geometric task and motion planning (GTAMP). GTAMP is a subclass of task and motion planning in which the goal is to move multiple objects to target regions among movable obstacles. A standard graph search algorithm is not directly applicable, because GTAMP problems involve hybrid search spaces and expensive action feasibility checks. To handle this, we introduce a novel planner that extends basic heuristic search with random sampling and a heuristic function that prioritizes feasibility checking on promising state action pairs. The main drawback of such pure planners is that they lack the ability to learn from planning experience to improve their efficiency. We propose two learning algorithms to address this. The first is an algorithm for learning a rank function that guides the discrete task level search, and the second is an algorithm for learning a sampler that guides the continuous motionlevel search. We propose design principles for designing data efficient algorithms for learning from planning experience and representations for effective generalization. We evaluate our framework in challenging GTAMP problems, and show that we can improve both planning and data efficiency