Multiple Agents Research Articles

Smart Microneedle Arrays Integrating Cell-Free Therapy and Nanocatalysis to Treat Liver Fibrosis.

Liver fibrosis is a chronic pathological condition lacking specific clinical treatments. Stem cells, with notable potential in regenerative medicine, offer promise in treating liver fibrosis. However, stem cell therapy is hindered by potential immunological rejection, carcinogenesis risk, efficacy variation, and high cost. Stem cell secretome-based cell-free therapy offers potential solutions to address these challenges, but it is limited by low delivery efficiency and rapid clearance. Herein, an innovative approach for in situ implantation of smart microneedle (MN) arrays enabling precisely controlled delivery of multiple therapeutic agents directly into fibrotic liver tissues is developed. By integrating cell-free and platinum-based nanocatalytic combination therapy, the MN arrays can deactivate hepatic stellate cells. Moreover, they promote excessive extracellular matrix degradation by more than 75%, approaching normal levels. Additionally, the smart MN arrays can provide hepatocyte protection while reducing inflammation levels by ≈70-90%. They can also exhibit remarkable capability in scavenging almost 100% of reactive oxygen species and alleviating hypoxia. Ultimately, this treatment strategy can effectively restrain fibrosis progression. The comprehensive in vitro and in vivo experiments, supplemented by proteome and transcriptome analyses, substantiate the effectiveness of the approach in treating liver fibrosis, holding immense promise for clinical applications.

Modeling coupled driving behavior during lane change: A multi-agent Transformer reinforcement learning approach

In a lane change (LC) scenario, the lane change vehicle interacts with surrounding vehicles. The interactions not only affect their driving behaviors but also influence the traffic flow. This study aims to model the coupled behavior of the lane changer and the follower in the target lane during LC. Large-scale real-world connected vehicle (CV) data from the Safety Pilot Model Deployment (SPMD) program are used to extract LCs and study vehicle interactions. A multi-agent Transformer-based deep deterministic policy gradient (MA-TDDPG) method is proposed to model the coupled behaviors during LC. The multi-agent framework can handle the multiple agents’ behaviors with interactions, and the Transformer can process the observation-action memory accurately and efficiently. The MA-TDDPG algorithm can learn the sequential decision-making process over continuous action space during LC with the accommodation of the multi-vehicle interaction and the driver’s memory effect. Compared to traditional supervised learning and reinforcement learning methods, it demonstrates superior performance in imitating the longitudinal and lateral actions of the lane changer and the follower. The findings of this study provide insights into the development of microscopic simulations by producing realistic LC behaviors, and assistance/automation LC systems by generating LC motions and responses conforming to human driving habits. The model also creates an interactive simulation environment and lays the foundation for optimizing driving strategies.

Investigating the causal effect of previously reported therapeutic agents for colorectal cancer prevention: protocol for a Mendelian randomization analysis.

Colorectal cancer (CRC) is the third most common cancer worldwide, with 1.9 million new cases in 2020 and a predicted rise to 3.2 million in 2040. Screening programmes are already in place to aid early detection and secondary prevention of CRC, but the rising prevalence means additional approaches are required in both primary and secondary prevention settings. Preventive therapy, whereby natural or synthetic agents are used to prevent, reverse or delay disease development, could be an effective strategy to further reduce cancer risk and potential agents have already been identified in conventional observational studies. However, as such studies are vulnerable to confounding and reverse causation, we aim to evaluate these observed relationships using Mendelian randomization (MR), an alternative causal inference approach which should be less susceptible to these biases. We will use two-sample MR, which uses two independent samples for the exposure and outcome data, to investigate previously reported observational associations of multiple potential preventive agents with CRC risk. We define preventive agents as any synthetic (e.g. approved medication) or natural (e.g. micronutrient, endogenous hormone) molecule used to reduce the risk of cancer. We will first extract potential preventive agents that have been previously linked to CRC risk in observational studies from reviews of the literature. We will then evaluate whether we can develop a genetic instrument for each preventive agent from previously published genome-wide association studies (GWASs) of direct measures of molecular traits (e.g. circulating levels of protein drug targets, blood-based biomarkers of dietary vitamins). The summary statistics from these GWASs, and a large GWAS of CRC, will be used in two-sample MR analyses to investigate the causal effect of putative preventive therapy agents on CRC risk. Sensitivity analyses will be conducted to evaluate the robustness of findings to potential violations of MR assumptions.

Achieving safe minimum circle circumnavigation around multiple targets: a dynamic compensation approach

Minimum circle circumnavigation is proposed in this paper, which is of special value in target monitoring, capturing and/or attacking. With this new pattern, our goal is to steer one agent to enclose multiple targets along a safe and minimum circular orbit with the desired tangential speed. This is achieved by a control protocol with dynamic compensators, which is also capable of avoiding collision between the agent and the multiple targets during the whole operational process. The control protocol is further extended to the scenarios of multiple agents circulating a group of targets by adding a coordination mechanism into the tangential velocity term, which allows the agents to distribute evenly on each expected circular orbit with the same or different radius. Simulations results illustrate the effectiveness of the proposed methods.

Synthesis and Verification of Mission Plans for Multiple Autonomous Agents under Complex Road Conditions

Mission planning for multi-agent autonomous systems aims to generate feasible and optimal mission plans that satisfy given requirements. In this article, we propose a tool-supported mission-planning methodology that combines (i) a path-planning algorithm for synthesizing path plans that are safe in environments with complex road conditions, and (ii) a task-scheduling method for synthesizing task plans that schedule the tasks in the right and fastest order, taking into account the planned paths. The task-scheduling method is based on model checking, which provides means of automatically generating task execution orders that satisfy the requirements and ensure the correctness and efficiency of the plans by construction. We implement our approach in a tool named MALTA, which offers a user-friendly GUI for configuring mission requirements, a module for path planning, an integration with the model checker UPPAAL, and functions for automatic generation of formal models, and parsing of the execution traces of models. Experiments with the tool demonstrate its applicability and performance in various configurations of an industrial case study of an autonomous quarry. We also show the adaptability of our tool by employing it in a special case of an industrial case study.

Self-collaboration Code Generation via ChatGPT

Although Large Language Models (LLMs) have demonstrated remarkable code-generation ability, they still struggle with complex tasks. In real-world software development, humans usually tackle complex tasks through collaborative teamwork, a strategy that significantly controls development complexity and enhances software quality. Inspired by this, we present a self-collaboration framework for code generation employing LLMs, exemplified by ChatGPT. Specifically, through role instructions, 1) Multiple LLM agents act as distinct ‘experts’, each responsible for a specific subtask within a complex task; 2) Specify the way to collaborate and interact, so that different roles form a virtual team to facilitate each other’s work, ultimately the virtual team addresses code generation tasks collaboratively without the need for human intervention. To effectively organize and manage this virtual team, we incorporate software-development methodology into the framework. Thus, we assemble an elementary team consisting of three LLM roles (i.e., analyst, coder, and tester) responsible for software development’s analysis, coding, and testing stages. We conduct comprehensive experiments on various code-generation benchmarks. Experimental results indicate that self-collaboration code generation relatively improves 29.9%-47.1% Pass@1 compared to the base LLM agent. Moreover, we showcase that self-collaboration could potentially enable LLMs to efficiently handle complex repository-level tasks that are not readily solved by the single LLM agent.

MET-MAPF: A Metamorphic Testing Approach for Multi-Agent Path Finding Algorithms

The Multi-Agent Path Finding (MAPF) problem, i.e., the scheduling of multiple agents to reach their destinations, has been widely investigated. Testing MAPF systems is challenging, due to the complexity and variety of scenarios and the agents’ distribution and interaction. Moreover, MAPF testing suffers from the oracle problem, i.e., it is not always clear whether a test shows a failure or not. Indeed, only considering whether the agents reach their destinations without collision is not sufficient. Other properties related to the “quality” of the generated paths should be assessed, e.g., an agent should not follow an unnecessarily long path. To tackle this issue, this paper proposes MET-MAPF, a Metamorphic Testing approach for MAPF systems. We identified ten Metamorphic Relations (MRs) that a MAPF system should guarantee, designed over the environment in which agents operate, the behaviour of the single agents, and the interactions among agents. Starting from the different MRs, MET-MAPF automatically generates test cases addressing them, so possibly exposing different types of failures. Experimental results show that MET-MAPF can indeed find MR violations not exposed by approaches that only consider the completion of the mission as test oracle. Moreover, experiments show that different MRs expose different types of violations.

An effective approach for reducing data redundancy in multi-agent system communication

The redundancy of the data is an active research topic. While an agent works in a multi-agent system, the number of messages between them increases. This is due to the fact that the functionalities data depends on other agents in terms of functional requirements. Typically, only one agent in a multi-agent system is responsible for accessing a database instead of replicating the database on each agent. A database is stored on multiple agents rather than a single agent to avoid a single point of failure. In this approach, the system has a higher load because one agent is responsible for all agent queries and must send duplicate messages to multiple agents, resulting in redundant data. In this research, we present Multi-Agent System for Commodity Data (MASCD) framework, the multi-agent system based communication using the distributed hash system, to reduce data redundancy in multi-agent system communication. Our anticipated method demonstrated how we divided the database names and efficiently distributed data to each agent. The database splitting is based on manufacturer names or product names. We utilize a table based on prime numbers. Through the hash function, we ascertain the index of the agent granted access to the relevant data. Each agent is accountable for its data. We use a Distributed Hash Table for efficient querying that stores data as key-value pairs. Each agent maintains a Finger Table containing the next and previous nodes for agent communication purposes. Using FIPA messages, we demonstrated how an agent could interact optimally. In conclusion, we illustrate the application of the proposed approach through a case study of mobile phones and university information systems.

Combination Foam and Liquid Sclerotherapy for Lower Extremity Reticular and Telangiectatic Veins: A Single-Center Retrospective Study.

Lower extremity reticular and telangiectatic veins are of common cosmetic concern. To retrospectively evaluate results of lower extremity sclerotherapy using a combination of foam and liquid sclerosing agents. A retrospective chart review of sclerotherapy patients at a dermatology practice (January 2014 to April 2023) was performed. Eight hundred and nine patients (775 women and 34 men) with a mean age of 49.6 ± 12.2 (18-84) years underwent a mean 1.4 ± 0.7 (1-7) sessions. Multiple different sclerosing agents were used, with 0.2% sodium tetradecyl sulfate foam/liquid predominating, although 72% glycerin liquid and 0.25% to 0.5% polidocanol foam/liquid were also used. Coagula occurred in 61.0% of patients at 2 weeks and 6.4% at 3 months, whereas postsclerotherapy hyperpigmentation was seen in 5.0% and 19.1% of patients at the same time points. Both were more common after first-round treatment, each with a trend toward decreased frequency with increasing session number. Telangiectatic matting was found in 2.3% of 3-month follow-up patients. Edema, superficial venous thrombophlebitis, migraines, and ulceration were rarely seen. Significant clinical improvement was noted in 72.0% of treatments. This retrospective chart review, the largest to date of its kind, confirms the safety and efficacy of cosmetic lower extremity sclerotherapy with a combination of foam and liquid sclerosing agents.

HyperComm: Hypergraph-based communication in multi-agent reinforcement learning

In the realm of fully cooperative multi-agent reinforcement learning (MARL), effective communication can induce implicit cooperation among agents and improve overall performance. In current communication strategies, agents are allowed to exchange local observations or latent embeddings, which can augment individual local policy inputs and mitigate uncertainty in local decision-making processes. Unfortunately, in previous communication schemes, agents may potentially receive irrelevant information, which increases training difficulty and leads to poor performance in complex settings. Furthermore, most existing works lack the consideration of the impact of small coalitions formed by agents in the multi-agent system. To address these challenges, we propose HyperComm, a novel framework that uses the hypergraph to model the multi-agent system, improving the accuracy and specificity of communication among agents. Our approach brings the concept of hypergraph for the first time in multi-agent communication for MARL. Within this framework, each agent can communicate more effectively with other agents within the same hyperedge, leading to better cooperation in environments with multiple agents. Compared to those state-of-the-art communication-based approaches, HyperComm demonstrates remarkable performance in scenarios involving a large number of agents.

Multi-Etiological Abortion due to Campylobacter spp. and Chlamydia abortus in a Sheep

In this case, multi-etiologic abortion due to Campylobacter spp. and Chlamydia abortus was determined in an aborted sheep fetus sent to Konya Veterinary Control Institute (KVKE) from a sheep farm operating in Aksaray province in 2020. The presence of Campylobacter spp. was determined by the bacterial isolation method. C.abortus was identified by qPCR and immunohistochemistry (IHC) methods. This study should be published in order to emphasize that in the fight against sheep abortions, multi-etiologic abortions involving multiple agents should be taken into consideration in addition to a single etiologic agent, and a multidisciplinary approach should be used in the diagnosis to identify the abortion agents.

Beyond Algorithmic Disclosure for Generative AI

One of the most commonly recommended policy interventions with respect to algorithms in general and artificial intelligence (“AI”) systems in particular is the need for greater transparency, often focusing on the disclosure of the variables employed by the algorithm and the weights given to those variables. This Essay argues that any meaningful transparency regime must provide information on other critical dimensions as well. For example, any transparency regime must also include key information about the data on which the algorithm was trained, including its source, scope, quality, and inner correlations, subject to constraints imposed by copyright, privacy, and cybersecurity law. Disclosures about pre-release testing also play a critical role in understanding an AI system’s robustness and its susceptibility to specification gaming. Finally, the fact that AI, like all complex systems, tends to exhibit emergent phenomena, such as proxy discrimination, interactions among multiple agents, the impact of adverse environments, and the well-known tendency of generative AI to hallucinate, makes ongoing post-release evaluation a critical component of any system of AI transparency.

Determinants of Course of Type 2 Diabetes in Indian Adolescents.

To identify determinants of the course of Type 2 diabetes in Indian adolescents. Records of 37 adolescents (24 boys; 29 post pubertal and eight pubertal) with Type 2 diabetes (initial HbA1C 10.1 ± 1.9% and BMI SDS 2.0 ± 0.8; family history of diabetes in 33, 89.2%) diagnosed at 15.2 ± 2.5 y and followed up for 3.8 ± 2.2 y till 19.1 ± 3.3 y of age, were reviewed. Initial treatment included insulin in 11 (29.7%), metformin alone in 22 (59.5%), and a combination of anti-diabetic medication in four (10.8%). Eleven subjects (29.7%) achieved remission at a median period of 5.3 mo (IQR- 17.13) after diagnosis; six of these relapsed within 0.9 ± 0.3 (range 0.4-1.3) y. The proportion of subjects requiring multiple anti-diabetic agents increased over follow-up (19% at six months, 32.5% at one year, 50% at two years, 59.1% at three, and 64.8% at four years), with the need for combination therapy after 0.9 ± 1.4 y. At the last follow-up, five were off treatment (13.5%), 10 (27%) were on metformin alone, and 22 (59.5%) were on multiple medications. The need for combination therapy at the last follow-up was lower in subjects with remission (27% against 73.1%, p = 0.02). The findings of this study suggest delayed presentation and rapid progression of Type 2 diabetes in Indian adolescents. Diagnosis on screening and achievement of remission were predictors of good outcome.

Radiosurgery to 5 or more newly diagnosed brain metastases in the systemic therapy era.

2026 Background: Stereotactic radiosurgery (SRS) is an ablative modality for focal treatment of brain metastases (BM) and is standard of care for limited BM. Suitability for SRS has historically been linked to number of BM, with up to 4 BM considered amenable to SRS. JLGK0901 revealed no difference in overall survival (OS) and similar rates of adverse events for patients with 5-10 vs. 2-4 BM. Evidence supporting for SRS for >10 BM is limited. As systemic treatments have evolved to include agents with improved central nervous system (CNS) efficacy and OS, the role of SRS warrants re-evaluation. Methods: We retrospectively reviewed patients from a single institution with ≥5 BM treated with SRS with no prior history of brain radiation 2015-2022 on an IRB-approved protocol. Clinical history, including sex, histology, Karnofsky performance status (KPS), extracranial disease status and systemic regimen used before and after diagnosis of BM, was reviewed. Systemic regimens were categorized based on potential for CNS efficacy as 1) none: no evidence, 2) weak: single agent immunotherapy or known but limited data on CNS efficacy, or 3) strong: multiple agents with known CNS efficacy, single agents with strong CNS efficacy. Kaplan-Meier method was used to estimate survival with log-rank test used to evaluate differences in survival curves. Cox proportional hazards modeling was used to evaluate differences in CNS progression-free survival (CNS-PFS) and OS. Results: 558 patients (48% female) who received SRS to ≥5 BM (range 5-23) were identified. Primary histologies included: 36% non-small cell lung cancer; 22% melanoma; 14% breast; 7% renal; 5% gastrointestinal; 15% other. Median OS was 11.0 months and did not differ between patients with 5-9 (n=441) vs ≥10 BM (n=117) (median OS 11.1 vs. 10.0 months; p=0.53). Median CNS-PFS was 6.6 months and did not differ between 5-9 vs ≥ 10 BM (median CNS-PFS 6.4 vs. 7.7 p=0.86). On univariate analysis, high KPS (HR 0.96; p<0.001) was associated with improved OS. Histology ( p=0.14), BM number ( p=0.53); average BM volume ( p=0.48), and systemic disease status ( p=0.17) were not associated with OS. Change of systemic therapy at time of SRS was associated with improved OS (HR 0.58; p<0.001) and CNS-PFS (HR 0.63; p<0.001). Amongst patients with systemic therapy change, regimens with increased CNS efficacy compared to prior therapy were associated with improved OS (HR 0.59; p<0.001) and CNS-PFS (HR 0.65; p=0.002). Conclusions: In this large, retrospective analysis of patients with ≥ 5 previously untreated BM treated with SRS, we found no association between BM number and CNS-PFS or OS. Change in systemic therapy, particularly to agents with CNS efficacy, was associated with improved outcomes. While limited by biases inherent to retrospective analyses, these results suggest patients with larger number of BM (≥10) are appropriate for SRS, especially when changing to systemic therapy with CNS penetrance.

Patient-reported treatment-related symptoms and use of the PRO-CTCAE for multiple myeloma clinical trials: A qualitative interview study.

e19516 Background: Therapeutic regimens for multiple myeloma (MM) involve multiple pharmacological agents, including steroids. Many cause treatment-related symptoms that add to patient burden or may trigger treatment cessation. This study gathered qualitative data about the patient experience with MM treatments, including regimens with steroids. Findings were summarized and compared to the Patient Reported Outcome version of the Common Terminology Criteria for Adverse Events (PRO-CTCAE), to assess its comprehensiveness and relevance in MM clinical trials. Methods: Individual, semi-structured interviews were conducted virtually with a diverse sample of 16 US-based, English-speaking adults (38% male, 25% Black, 13% Hispanic) to explore their experiences with MM and treatments, including perspectives on steroids and potential benefits of steroid-sparing regimens. Content analysis was used to summarize and compare terminology participants used to describe their experiences with treatments. Results: Participants reported a diverse set of treatment-related symptoms impacting their overall experience, indicating the importance of assessing regimen tolerability in clinical trials. Steroids were linked to increased energy that interfered with sleep or functioning, irritability, taste changes, weight gain, increased appetite, and cognitive symptoms like brain fog. Participants reported that most symptoms were tolerable, but they would prefer to avoid steroids if treatment benefits did not diminish. Many reported symptoms were included in the PRO-CTCAE items (see Table), but several symptoms related to regimens with steroids are not covered: excess energy, bitter/metallic taste, taste sensitivity, brittle nails, irritability, hunger/increased appetite, and muscle weakness. Conclusions: MM patients experience a range of treatment-related symptoms, many of which impact their quality of life. If given the option, patients would prefer a steroid-free regimen if there was proven clinical benefit. While the PRO-CTCAE covers most of the treatment-related symptoms, proving to be an option for multiple myeloma clinical trials, additional PRO instruments may be needed to cover specific gaps in the PRO-CTCAE item library content coverage. [Table: see text]

Множественная лекарственная устойчивость уропатогенов как фактор риска при оказании медицинской помощи (опыт Центральной Болгарии)

The emergence of resistance to multiple antimicrobial agents in uropathogenic bacteria has become a significant public health problem. These microorganisms are the cause of urinary tract infections, which are among the most frequently treated diseases in the outpatient medical care. A cross-sectional study was conducted among 353 outpatients over 18 years of age with positive urine sample culture. It involved measuring the index of multiple antibiotic resistance of uropathogens causing infections among individuals. The research was conducted in the period January – June 2023 at a private clinical laboratory center located in Stara Zagora, Bulgaria. One third of the isolates of E. coli and more than a half of those of Klebsiella pneumoniae had a measured MAR index greater than 0.2, with confirmatory high rates of resistance to aminopenicillins and cephalosporins. Alarmingly high MARI values were measured in all representatives of Enterococcus faecalis, Enterobacter species and Pseudomonas aeruginosa. These strains were also established to have high levels of resistance (varying from 58.3 to 100 %) to some of antimicrobial medicines most commonly prescribed in the outpatient healthcare. Findings of this study emphasize the importance and seriousness of the existing problem with the increasing prevalence of uropathogenic microorganisms with multiple resistance to oral antibiotics. This creates elevated risks of failure to provide effective treatment for patients.

A Case of Extra-Pulmonary Mycoplasma Pneumoniae Infection

Abstract Mycoplasma pneumoniae is a common atypical pathogen causing comunitary-acquired pneumonia (CAP), mostly in children and young adults with age between 5 and 40 years old. Beyond respiratory symptoms, extra-pulmonary features are described, such as skin rash, arthralgia, gastrointestinal symptoms, and heart disease, occurring in less than 5% to 10% of patients(1). This case refers to a 26-year-old patient, without significant medical history or background treatment, immunocompetent, who is presenting to the emergency room for prolonged fever, and conjunctival suffusions. Following the investigations, pneumonia, and conjunctivitis due to atypical bacteria are suspected and appropriate treatment is administered. Along with other multiple infectious agents that could precipitate a respiratory febrile infection with ocular involvement, we should consider also Mycoplasma pneumoniae as a possible cause. Appropriate antibiotic therapy may result in a favorable, but delayed course of the disease.

Long-term outcomes of noninfectious uveitis treated with systemic immunomodulatory therapy: a retrospective case series

ObjectiveTo study the clinical characteristics and long-term outcomes of patients with noninfectious uveitis (NIU) who are treated with systemic immunomodulatory therapy (IMT). DesignRetrospective case series. ParticipantsAll consecutive cases of adults with NIU under the care of 5 uveitis subspecialty tertiary care clinics between 2010 to 2021 were included. MethodsPatient outcomes were assessed at initial presentation and at the latest available follow-up. ResultsA total of 418 NIU patients receiving IMT therapy with a median age of 46.0 years and 59.3% female were identified. Each patient required an average of 1.4 agents until achieving an optimal response. Following initial treatment with prednisone, patients were most commonly initiated on methotrexate. The top 3 treatments with the highest proportion of optimal treatment response when taken alone or in combination with other agents were infliximab (79.3%), cyclosporine (75%), and adalimumab (70%). The strongest predictors for requiring a greater number of IMTs trialed were younger age, panuveitis, and a chronic or recurrent disease course. Multivariable linear regression analysis suggested that baseline visual acuity at diagnosis was the only significant predictor of final visual acuity (p < 0.001). ConclusionsNIU patients on IMT are often trialed on multiple therapeutic agents before achieving an optimal treatment response. Visual acuity at diagnosis is a predictor of final visual outcomes, whereas chronic or recurrent disease course, younger age, and panuveitis are predictors of requiring multiagent treatment regimens.

An Adversarial Approach for Automated Pokémon Team Building and Metagame Balance

Meta-game balance is a crucial task in game development, and automation of this process could assist game developers by vastly reducing time costs. We explore and evaluate a meta-game balance model over the recently proposed VGC AI Competition Framework. We propose an adversarial model where team builder agents try to maximize their win rate by narrowing to the most optimal team configurations, resulting in a reduction of the diversity of Pokémon employed, while a balancing agent re-adapts the Pokémon inner attributes to incentive the team builder agents to incorporate a greater variety of Pokémon into their teams increasing the meta-game's overall diversity and balance. one Furthermore, we developed multiple team builder agents divided into two groups: the first group assumes that individual Pokémon advantages are the primary factor to determine the outcome of game matches; the second group also exploits the implicit synergy between teammates. These agents make use of meta-gaming, linear optimization, and evolutionary search to find strong combinations against the current meta-game. The strongest team builder is faced against the team meta-game balance agent for its evaluation. Deep learning is also employed to predict the outcome of matches and recommend constructive elements of teams.

Decentralized PI-control and anti-windup in resource sharing networks

We consider control of multiple stable first-order agents which have a control coupling described by an M-matrix. These agents are subject to incremental sector-bounded input nonlinearities. We show that such plants can be globally asymptotically stabilized to a unique equilibrium using fully decentralized proportional–integral controllers equipped with anti-windup and subject to local tuning rules. In addition, we show that when the nonlinearities correspond to the saturation function, the closed loop asymptotically minimizes a weighted 1-norm of the agents state mismatch. The control strategy is finally compared to other state-of-the-art controllers on a numerical district heating example.