Agent Programming Research Articles

Dose rounding bevacizumab and its biosimilars to reduce drug waste in the oncology care model in a community oncology network.

1529 Background: The Oncology Care Model (OCM) was a 6-year long Medicare value-based care (VBC) program that rewarded practices for decreasing TCOC compared to a benchmark price, while maintaining high-quality cancer care. Drug waste from partially used single dose vials (SDV) of expensive anti-neoplastic agents is a leading contributor to avoidable Total Cost of Care (TCOC) expenditures for payers. Dose rounding (DR) of biological anti-neoplastic agents, using recommendations (https://www.nccn.org/docs/default-source/clinical/order-templates/hopa.pdf?sfvrsn=3af91118_6) from the Hematology Oncology Pharmacists Association (HOPA) and endorsed by the National Comprehensive Cancer Network (NCCN), is a standard approach to mitigate the impact of drug waste. Many of the practices in The US Oncology Network (The Network) adopted a DR program for biologic agents in 2018 as a strategy to reduce drug waste and TCOC. We studied the impact of a DR initiative targeting bevacizumab and its biosimilars (B+B) on TCOC and drug waste expenditures for The Network practices participating in the OCM. Methods: Claims data for 14 practices in The Network participating in the OCM were assessed. Drug administration data for B+B was used to evaluate drug waste, total dose, TCOC, DR and the financial impact of drug waste reduction on TCOC from 2017 to 2021. Results: In 2017, prior to implementing DR, an average of 5.3% (range 4.6% to 5.5%) of the total dose of B+B was wasted, with ~25% of B+B administrations having waste in excess of 10% of the total dose. Implementation of the initiative led to a reduction of waste on 30% of the total administrations of B+B across 14 OCM participant practices in The Network. As a result of DR, B+B drug waste decreased by 55% to an average of 2.5% (range - 2.3% to 2.7%) of the total dose in 2021. Decreasing drug waste resulted in TCOC reduction of approximately $100 per member episode per month (0.97% of TCOC) for episodes with B+B administrations in the OCM. The highest waste reduction was seen with B+B administered for treatment of gastrointestinal cancers (colon, rectal, anal, gastroesophageal, or pancreatic cancers) compared to other cancer types (lung, brain, ovarian, etc.), and for B+B administrations where drug waste was greater than 10% of the total dose (~50% reduction). Conclusions: Drug waste contributes avoidable expenditures to the TCOC. Implementing strategies such as DR can offset the impact of drug waste and offer meaningful TCOC savings for VBC programs such as the OCM. A similar DR approach may be adopted for other high-cost biologic agents where drug waste results from partial use of SDVs.

Review of Restricted Experiment Requests, Division of Select Agents and Toxins, US Centers for Disease Control and Prevention, 2014-2021.

The US Centers for Disease Control and Prevention Division of Select Agents and Toxins (DSAT) regulates laboratories that possess, use, or transfer select agents and toxins within United States as part of the Federal Select Agent Program. DSAT also mitigates biosafety risks through the review of "restricted experiments," which under the select agent regulations are experiments that pose heightened biosafety risks. In a previous study, we evaluated restricted experimental requests submitted to DSAT for review between 2006 and 2013. The purpose of this study is to provide an updated analysis of requests to conduct potential restricted experiments submitted to DSAT between 2014 and 2021. This article describes the trends and characteristics of the data associated with restricted experimental requests involving select agents and toxins that have an impact on public health and safety (US Department of Health and Human Services agents only) or both public health and safety and animal health or products (overlap agents). From January 2014 to December 2021, DSAT received 113 requests to conduct potential restricted experiments; however, 82% (n=93) of those requests were determined not to meet the regulatory definition of a restricted experiment. Of the 20 requests that met the definition of a restricted experiment, 8 were denied because the experiments had the potential to compromise disease control in humans. DSAT continues to encourage entities to practice due diligence and request a review of research that could potentially meet the regulatory definition of a restricted experiment out of an abundance of caution to protect public health and safety and prevent any potential compliance action.

A formal semantics for agent distribution and fault tolerance in Jason

This article provides a formal specification of the distribution and fault-tolerance mechanisms of eJason. The eJason programming language is an extension to the agent-oriented programming language Jason that introduces native support for the transparent distribution of agents as well as fault-tolerance mechanisms. This formal semantics is presented from a multiagent system perspective. It unambiguously describes both the possible evolution of the distributed multiagent system over time and the different instruments for fault detection and fault recovery, hence exposing their strengths. This specification may serve as a reference for researchers interested in the inclusion of similar mechanisms in agent-oriented programming languages. The formal semantics has been mechanized through an (open-source) implementation written in Prolog, which implements both the standard Jason operational semantics, along with the new rules for distribution and fault-tolerance introduced in this article.

Deep reinforcement learning for turbulent drag reduction in channel flows

We introduce a reinforcement learning (RL) environment to design and benchmark control strategies aimed at reducing drag in turbulent fluid flows enclosed in a channel. The environment provides a framework for computationally efficient, parallelized, high-fidelity fluid simulations, ready to interface with established RL agent programming interfaces. This allows for both testing existing deep reinforcement learning (DRL) algorithms against a challenging task, and advancing our knowledge of a complex, turbulent physical system that has been a major topic of research for over two centuries, and remains, even today, the subject of many unanswered questions. The control is applied in the form of blowing and suction at the wall, while the observable state is configurable, allowing to choose different variables such as velocity and pressure, in different locations of the domain. Given the complex nonlinear nature of turbulent flows, the control strategies proposed so far in the literature are physically grounded, but too simple. DRL, by contrast, enables leveraging the high-dimensional data that can be sampled from flow simulations to design advanced control strategies. In an effort to establish a benchmark for testing data-driven control strategies, we compare opposition control, a state-of-the-art turbulence-control strategy from the literature, and a commonly used DRL algorithm, deep deterministic policy gradient. Our results show that DRL leads to 43% and 30% drag reduction in a minimal and a larger channel (at a friction Reynolds number of 180), respectively, outperforming the classical opposition control by around 20 and 10 percentage points, respectively.

IoT-Based Integrated Marketing Communication in Closing Quality Agent Marketing Program

The researcher examines a digital economy-based manufacturing industry company where the company has a Closing Quality Agent program that aims to screen the network below, namely screening Closing Quality Agent program that aims to filter the network under it, namely screening qualified and serious agents in running this business. The applicable Closing Quality Agent will have an impact on the entire network if the agent does not pass the qualification. This research was conducted to try to help and evaluate existing programs in the company The method for analyzing this research is qualitative by direct observation, in-depth interviews, collection of documentation, and secondary data of the company. Research is carried out in the company through the marketing division. Based on research that has been carried out, this program is less effective in maximizing its digital system. This research resulted in a program evaluation using the theory of Integrategrated Marketing Communication with the implementation of the Internet of Things which is expected to be implemented in the company system in supporting the Closing Quality Agent program is carried out.

Improving U.S. Biosafety and Biosecurity: Revisiting Recommendations from the Federal Experts Security Advisory Panel and the Fast Track Action Committee on Select Agent Regulations.

In response to a series of biosafety incidents in 2014, the White House directed two high-level expert committees to analyze biosafety and biosecurity in U.S. laboratories and make recommendations for work with select agents and toxins. Overall, they recommended 33 actions to address areas related to national biosafety, including promoting a culture of responsibility, oversight, outreach and education, applied biosafety research, incident reporting, material accountability, inspection processes, regulations and guidelines, and determining the necessary number of high-containment laboratories in the United States. The recommendations were collected and grouped into categories previously defined by the Federal Experts Security Advisory Panel and the Fast Track Action Committee. Open-source materials were examined to determine what actions had been taken to address the recommendations. The actions taken were compared against the reasoning provided in the committee reports to determine if the concerns were sufficiently addressed. In this study, we found that 6 recommendations were not addressed and 11 were insufficiently addressed out of 33 total recommended actions. Further work is needed to strengthen biosafety and biosecurity in U.S. laboratories handling regulated pathogens (biological select agents and toxins [BSAT]). These carefully considered recommendations should now be enacted, including determining if there is sufficient high-containment laboratory space for response to a future pandemic, developing a sustained applied biosafety research program to improve our understanding of how high-containment research should be performed, bioethics training to educate the regulated community on the consequences of unsafe practices in BSAT research, and the creation of a no-fault incident reporting system for biological incidents, which may inform and improve biosafety training. The work presented in this study is significant because previous incidents that occurred in Federal laboratories highlighted shortcomings in the Federal Select Agent Program and the Select Agent Regulations. Progress was made on implementing recommendations to address the shortcomings, but efforts were lost or forgotten over time. The COVID-19 pandemic has provided a brief window of interest in biosafety and biosecurity, and an opportunity to address these shortcomings to increase readiness for future disease emergencies.

Preparation of noninfectious scRNAseq samples from SARS-CoV-2-infected epithelial cells.

Coronavirus disease (COVID-19) is an infectious disease caused by the SARS coronavirus 2 (SARS-CoV-2) virus. Direct assessment, detection, and quantitative analysis using high throughput methods like single-cell RNA sequencing (scRNAseq) is imperative to understanding the host response to SARS-CoV-2. One barrier to studying SARS-CoV-2 in the laboratory setting is the requirement to process virus-infected cell cultures, and potentially infectious materials derived therefrom, under Biosafety Level 3 (BSL-3) containment. However, there are only 190 BSL3 laboratory facilities registered with the U.S. Federal Select Agent Program, as of 2020, and only a subset of these are outfitted with the equipment needed to perform high-throughput molecular assays. Here, we describe a method for preparing non-hazardous RNA samples from SARS-CoV-2 infected cells, that enables scRNAseq analyses to be conducted safely in a BSL2 facility-thereby making molecular assays of SARS-CoV-2 cells accessible to a much larger community of researchers. Briefly, we infected African green monkey kidney epithelial cells (Vero-E6) with SARS-CoV-2 for 96 hours, trypsin-dissociated the cells, and inactivated them with methanol-acetone in a single-cell suspension. Fixed cells were tested for the presence of infectious SARS-CoV-2 virions using the Tissue Culture Infectious Dose Assay (TCID50), and also tested for viability using flow cytometry. We then tested the dissociation and methanol-acetone inactivation method on primary human lung epithelial cells that had been differentiated on an air-liquid interface. Finally, we performed scRNAseq quality control analysis on the resulting cell populations to evaluate the effects of our virus inactivation and sample preparation protocol on the quality of the cDNA produced. We found that methanol-acetone inactivated SARS-CoV-2, fixed the lung epithelial cells, and could be used to obtain noninfectious, high-quality cDNA libraries. This methodology makes investigating SARS-CoV-2, and related high-containment RNA viruses at a single-cell level more accessible to an expanded community of researchers.

Simulation of passenger movement dynamics in a vehicle

Support The article deals with the problem of modeling the movement of passengers in the interior of a vehicle using the example of an airport bus. Such modeling will allow automating the process of optimizing the layout of the salon, provide comfortable conditions for passengers and improve the fullness of the salon. When modeling the dynamics of passenger movement, the paradigm of agen-toriented programming, based on the modeling of social intelligence, was used. A class of extremely simplified discrete models - cellular automata - was used here. Developed methods that take into account the behavior of large groups of people were also used. To implement the developed methodology, the version of airport buses used in the 50s and 60s of the last century - a truck with a passenger semi-trailer was - taken as an example.

XPlaM: A toolkit for automating the acquisition of BDI agent-based Digital Twins of organizations

A Digital Twin ideally manifests the same behaviour (in silico) of its physical counterpart. While considerable attention has been paid to the development of Digital Twins for physical devices/systems, the question of developing Twins for organizations has received relatively little attention. The setting in which we address this problem is very general and, consequently, very challenging. We look at the automatic acquisition of Digital Twins of organizations. To that end, this paper builds on the following two premises: (1) that Digital Twins of organizations can provide value, and (2) that Belief-Desire-Intention (BDI) agents are a particularly effective means for representing Digital Twins. The overall approach is to leverage the externally observable behaviour of the target system and then generate candidate BDI agent programs that best explain (in the sense of formal abduction) the observed behaviour. The candidate agent programs are generated by searching through potentially large hypotheses spaces for possible plans, selection functions and beliefs. The resulting approach suggests that using abduction to generate Digital Twins of organizations in the form of BDI agents can be effective.

Peningkatan Layanan Melalui Strategi Marketing Door To Door

Sharia Financing Savings and Loans Cooperative (KSPPS) BMT Al Yaman Wringinputih Muncar Banyuwangi is one of the institutions operating in the economic sector which has been located at the location of the Manbaul Ulum Banyuwangi Islamic Boarding School Foundation since 2016 with the slogan "Maintaining Trust, Achieving Barokah". KSPPS BMT Al Yaman Wringinputih Muncar Banyuwangi is a cooperative that organizes sharia savings and loan activities. KSPPS BMT Al Yaman in its marketing uses a Door to Door marketing strategy to attract customers. The aim of this research is to find out the Door to Door marketing strategy that has been carried out by KSPPS BMT Al Yaman Wringinputih Muncar Banyuwangi in improving KSPPS BMT Al Yaman Wringinputih transaction services and to find out the obstacles and efforts implemented. This research uses a field research method with a qualitative descriptive approach with data collection in the form of observation, indirect interviews and documentation taken from one hundred Wringinputih Muncar Banyuwangi customers. The results of this research are a Door to Door marketing strategy to improve services that have been carried out by KSPPS BMT Al Yaman, realizing that customers feel very helped, which creates a sense of trust in the services provided. However, there were several obstacles that emerged from KSPPS BMT Al Yeman which were followed up directly. From the efforts that have been made, there are several obstacles that must be optimized in improving services. Several suggestions that researchers gave to KSPPS BMT Al Yeman to optimize the Sharia Agent program, M-BMT Al Yaman, and public relations. Another suggestion is to provide guidance to the AO (Account Officer) to be confident in conveying cooperative product information to customers in order to gain customer loyalty

Modularization in Belief-Desire-Intention agent programming and artifact-based environments.

This article proposes an extension for the Agents and Artifacts meta-model to enable modularization. We adopt the Belief-Desire-Intention (BDI) model of agency to represent independent and reusable units of code by means of modules. The key idea behind our proposal is to take advantage of the syntactic notion of namespace, i.e., a unique symbol identifier to organize a set of programming elements. On this basis, agents can decide in BDI terms which beliefs, goals, events, percepts and actions will be independently handled by a particular module. The practical feasibility of this approach is demonstrated by developing an auction scenario, where source code enhances scores of coupling, cohesion and complexity metrics, when compared against a non-modular version of the scenario. Our solution allows to address the name-collision issue, provides a use interface for modules that follows the information hiding principle, and promotes software engineering principles related to modularization such as reusability, extensibility and maintainability. Differently from others, our solution allows to encapsulate environment components into modules as it remains independent from a particular BDI agent-oriented programming language.

Event-Triggered Observers and Distributed H∞ Control of Physically Interconnected Nonholonomic Mechanical Agents in Harsh Conditions

In this article, event-triggered (ET) observers and ET-distributed <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${\boldsymbol {\mathcal {H}}_{\boldsymbol \infty }}$ </tex-math></inline-formula> controllers are investigated for physically interconnected nonholonomic mechanical agents in harsh conditions, such as skidding, slipping, and dead-zone disturbances. The agents’ models are presented by strict-feedback nonlinear large-scale systems, but unlike the existing studies, unknown dynamics with feedback outputs are considered, and assumptions of polynomial-type nonlinearities for physical interconnection functions are relaxed. Initially, the observer of unmeasurable states via the outputs is designed. Then, ET augmented controllers are established to transform the physically interconnected system into isolated subsystems connected by a communication network. By utilizing the local information of each agent and adaptive dynamic programming (ADP), ET-distributed <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">${\boldsymbol {\mathcal {H}}_{\boldsymbol \infty }}$ </tex-math></inline-formula> optimal control laws and disturbance rejection laws are derived. The parameters of the ET observers and controllers are synchronously updated online by event-triggering mechanisms; thus, it reduces computational complexity and communication. It is guaranteed that the states in the closed dynamics are <inline-formula xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> <tex-math notation="LaTeX">$L_{2}$ </tex-math></inline-formula> -gain bounded and the Zeno behavior is avoided. Additionally, the convergence of cost functions to the near-optimal values is accelerated by a concurrent learning technique, avoiding dependence on the online examination of the persistence of excitation conditions. Finally, the control performance of a nonholonomic multirobot system in a comparative simulation study shows that the proposed observers and controllers are effective.

Accountability in multi-agent organizations: from conceptual design to agent programming

Accountability in multi-agent organizations: from conceptual design to agent programming

Content Analysis of Agricultural Extension Agent Program on Misr Agricultural Channel During the Period From 1/6/2021 To 31/12/2021

Content Analysis of Agricultural Extension Agent Program on Misr Agricultural Channel During the Period From 1/6/2021 To 31/12/2021

Collaborative Autonomy: Human–Robot Interaction to the Test of Intelligent Help

A big challenge in human–robot interaction (HRI) is the design of autonomous robots that collaborate effectively with humans, exposing behaviors similar to those exhibited by humans when they interact with each other. Indeed, robots are part of daily life in multiple environments (i.e., cultural heritage sites, hospitals, offices, touristic scenarios and so on). In these contexts, robots have to coexist and interact with a wide spectrum of users not necessarily able or willing to adapt their interaction level to the kind requested by a machine: the users need to deal with artificial systems whose behaviors must be adapted as much as possible to the goals/needs of the users themselves, or more in general, to their mental states (beliefs, goals, plans and so on). In this paper, we introduce a cognitive architecture for adaptive and transparent human–robot interaction. The architecture allows a social robot to dynamically adjust its level of collaborative autonomy by restricting or expanding a delegated task on the basis of several context factors such as the mental states attributed to the human users involved in the interaction. This collaboration has to be based on different cognitive capabilities of the robot, i.e., the ability to build a user’s profile, to have a Theory of Mind of the user in terms of mental states attribution, to build a complex model of the context, intended both as a set of physical constraints and constraints due to the presence of other agents, with their own mental states. Based on the defined cognitive architecture and on the model of task delegation theorized by Castelfranchi and Falcone, the robot’s behavior is explainable by considering the abilities to attribute specific mental states to the user, the context in which it operates and its attitudes in adapting the level of autonomy to the user’s mental states and the context itself. The architecture has been implemented by exploiting the well known agent-oriented programming framework Jason. We provide the results of an HRI pilot study in which we recruited 26 real participants that have interacted with the humanoid robot Nao, widely used in HRI scenarios. The robot played the role of a museum assistant with the main goal to provide the user the most suitable museum exhibition to visit.

DTWN: Q-learning-based Transmit Power Control for Digital Twin WiFi Networks

Interference has always been the main threat to the performance of traditional WiFi networks and next-generation moving forward. The problem can be solved with transmit power control(TPC). However, to accomplish this, an information-gathering process is required. But this brings overhead concerns that decrease the throughput. Moreover, mitigation of interference relies on the selection of transmit powers. In other words, the control scheme should select the optimum configuration relative to other possibilities based on the total interference, and this requires an extensive search. Furthermore, bidirectional communication in real-time needs to exist to control the transmit powers based on the current situation. Based on these challenges, we propose a complete solution with Digital Twin WiFi Networks (DTWN). Contrarily to other studies, with the agent programs installed on the APs in the physical layer of this architecture, we enable information-gathering without causing overhead to the wireless medium. Additionally, we employ Q-learning-based TPC in the Brain Layer to find the best configuration given the current situation. Consequently, we accomplish real-time monitoring and management thanks to the digital twin. Then, we evaluate the performance of the proposed approach through total interference and throughput metrics over the increasing number of users. Furthermore, we show that the proposed DTWN model outperforms existing schemes.

Application of Multi-Criteria Decision Analysis Techniques for Informing Select Agent Designation and Decision Making.

The Centers for Disease Control and Prevention (CDC) Select Agent Program establishes a list of biological agents and toxins that potentially threaten public health and safety, the procedures governing the possession, utilization, and transfer of those agents, and training requirements for entities working with them. Every 2 years the Program reviews the select agent list, utilizing subject matter expert (SME) assessments to rank the agents. In this study, we explore the applicability of multi-criteria decision analysis (MCDA) techniques and logic tree analysis to support the CDC Select Agent Program biennial review process, applying the approach broadly to include non-select agents to evaluate its generality. We conducted a literature search for over 70 pathogens against 15 criteria for assessing public health and bioterrorism risk and documented the findings for archiving. The most prominent data gaps were found for aerosol stability and human infectious dose by inhalation and ingestion routes. Technical review of published data and associated scoring recommendations by pathogen-specific SMEs was found to be critical for accuracy, particularly for pathogens with very few known cases, or where proxy data (e.g., from animal models or similar organisms) were used to address data gaps. Analysis of results obtained from a two-dimensional plot of weighted scores for difficulty of attack (i.e., exposure and production criteria) vs. consequences of an attack (i.e., consequence and mitigation criteria) provided greater fidelity for understanding agent placement compared to a 1-to-n ranking and was used to define a region in the upper right-hand quadrant for identifying pathogens for consideration as select agents. A sensitivity analysis varied the numerical weights attributed to various properties of the pathogens to identify potential quantitative (x and y) thresholds for classifying select agents. The results indicate while there is some clustering of agent scores to suggest thresholds, there are still pathogens that score close to any threshold, suggesting that thresholding “by eye” may not be sufficient. The sensitivity analysis indicates quantitative thresholds are plausible, and there is good agreement of the analytical results with select agent designations. A second analytical approach that applied the data using a logic tree format to rule out pathogens for consideration as select agents arrived at similar conclusions.

Modelling and verifying BDI agents with bigraphs

• We show BDI agent programming languages can be encoded in Milner's Bigraphs. • We prove the encoding of a superset of AgentSpeak, e.g. CAN language, is faithful. • We provide a verified executable semantics for CAN language programs. • We verify CAN language program properties through model checking. • This is a foundation to enable advanced reasoning, e.g. probabilistic verification. The Belief-Desire-Intention (BDI) architecture is a popular framework for rational agents; existing verification approaches either directly encode simplified (e.g. lacking features like failure recovery) BDI languages into existing verification frameworks (e.g. Promela), or reason about specific BDI language implementations . We take an alternative approach and employ Milner's bigraphs as a modelling framework for a fully featured BDI language, the Conceptual Agent Notation (CAN)—a superset of AgentSpeak featuring declarative goals, concurrency, and failure recovery. We provide an encoding of the syntax and semantics of Can agents, and give a rigorous proof that the encoding is faithful. Verification is based on the use of mainstream software tools including BigraphER, and a small case study verifying several properties of Unmanned Aerial Vehicles (UAVs) illustrates the framework in action. The executable framework is a foundational step that will enable more advanced reasoning such as plan preference, intention priorities and trade-offs, and interactions with an environment under uncertainty.

Business models and sustainability: opportunities and challenges perceived by micro and small Brazilian companies

This article aims to identify opportunities and challenges of innovation for sustainability perceived by micro and small enterprises (MSEs) for developing new business models. Questionnaires were applied to companies participating in the local innovation agent (agente local de inovação - ALI) Program in a Brazilian city to survey their perceptions and management practices concerning sustainability and innovation that allow the development of new business models, with 13 responses being obtained. Descriptive statistics analysed the data. The results show that innovation for sustainability generates opportunities such as new businesses, strengthening the supplier network, winning over new customers and generating better results. The study also highlighted challenges in business model innovation for sustainability, such as: finding suppliers and partners for sustainable inputs that practice sustainability, information and incentives from the public authorities, customer culture for acquiring sustainable products and services and lack of financial resources to invest in innovation were found.

Smallholders perception on the roles of extension agent - coconut industry sustainability in Tanjong Karang, Selangor, Malaysia

This study was conducted to analyse the smallholder’s perception of the roles of extension agents - coconut industry sustainability. To ensure the extension agent program is successful, the agent should recognize their roles as the change agent in the coconut industry which is, they act as a catalyst, process helpers, resource linkers, and solution giver. This study was conducted in sub-districts of Kuala Selangor at Tanjong Karang. The sample of fifty-six respondents was selected from the 65 population through the random sampling method. By using a survey method, the questionnaires were distributed to measure the roles of the extension agent as a change agent based on the smallholder’s perception. Based on correlation coefficient analysis, it was shown that all data was significant between catalyst, process helper, resource linker, and solution giver by using 0.10 (p-value). The highest correlation coefficient in this analysis is solution giver that shows the correlation value 0.660 and can best be described as a positive strong relationship. The result for multiple linear regression showed that process helper was the dominant factor for this study which showed the beta, p-value is 0.264 by using 0.10 (pvalue). Hence, it can be analysed that the level of the extension agent roles in the coconut industry is at a medium level because they are not fully utilizing their roles as a change agent. As for the recommendation, extension agents in the coconut industry should be more aggressive by looking at the various aspects such as communication, decision making, knowledge, skills, and others to make the coconut industry sustainable for the future.