Safe Action Research Articles

Reduction of Home Deliveries in Kaoma District: Impact of Safe Motherhood Action Groups from 2020 to 2021

Reduction of Home Deliveries in Kaoma District: Impact of Safe Motherhood Action Groups from 2020 to 2021

Trends of Antenatal Care Visits in Kaoma District after Training of Safe Motherhood Action Groups from 2020 to 2021

Trends of Antenatal Care Visits in Kaoma District after Training of Safe Motherhood Action Groups from 2020 to 2021

Oral nutritional supplementation as an intervention in patients with hip fracture

Fragility hip fracture (FHR) is an increasingly prevalent pathology in industrialized countries, with high social and health costs. Malnutrition or the risk of malnutrition in the population with FHR is too common and negatively impacts the clinical course, the rehabilitation process, and the resulting quality of life after discharge. Technological advances applied to the diagnosis of the nutritional status of patients with FHR contribute to achieving an early intervention of malnutrition and reducing associated morbidity and mortality. Oral nutritional supplementation with a high energy and protein intake could be a safe and effective action to improve the prognosis of patients with FHR, but further research is required to confirm this.

The efficacy of picture vs. word danger cues in reducing imitation of dangerous actions in children.

Children are naturally curious and often have limited self-control, leading them to imitate both safe and dangerous actions. This study aimed to investigate whether dangerous cues could effectively inhibit children's imitation of hazardous behaviors and to compare the effectiveness of picture cues versus word cues in reducing this imitation. Seventy-six children were divided into two groups: one group received picture cues, and the other received word cues. Both groups observed an agent grasping an object and were instructed to perform a corresponding keystroke response when a number appeared. A comparable group of adults was also included for reference. The results demonstrated that picture cues were significantly more effective than word cues in reducing the children's tendency to imitate dangerous actions. These findings suggest that picture cues are a more effective method for preventing imitation of risky behaviors in children, which has important implications for improving safety education and accident prevention strategies through the use of visual danger cues.

A train trajectory optimization method based on the safety reinforcement learning with a relaxed dynamic reward

Train trajectory optimization (TTO) is an effective way to address energy consumption in rail transit. Reinforcement learning (RL), an excellent optimization method, has been used to solve TTO problems. Although traditional RL algorithms use penalty functions to restrict the random exploration behavior of agents, they cannot fully guarantee the safety of the process and results. This paper proposes a proximal policy optimization based safety reinforcement learning framework (S-PPO) for the train trajectory optimization, including a safe action rechoosing mechanism (SARM) and a relaxed dynamic reward mechanism (RDRM) combining a relaxed sparse reward and a dynamic dense reward. SARM guarantees that the new states generated by the agent consistently adhere to the environmental security constraints, thereby enhancing sampling efficiency and facilitating algorithm convergence. RDRM makes it easier for agents to obtain successful samples by relaxing time constraints, which also offers a better balance between exploration and exploitation. The experimental results show that S-PPO can significantly improve performance and obtain better train operation trajectories than soft constraint methods, and the convergence process is smoother. Finally, it was demonstrated that S-PPO exhibits good adaptability across various speed limit tracks.

Nanoemulsions of essential oils against multi-resistant microorganisms: An integrative review

Microbial resistance to drugs continues to be a global public health issue that demands substantial investment in research and development of new antimicrobial agents. Essential oils (EO) have demonstrated satisfactory and safe antimicrobial action, being used in pharmaceutical, cosmetic, and food formulations. In order to improve solubility, availability, and biological action, EO have been converted into nanoemulsions (NE). This review identified scientific evidence corroborating the antimicrobial action of nanoemulsions of essential oils (NEEO) against antibiotic-resistant pathogens. Using integrative review methodology, eleven scientific articles evaluating the antibacterial or antifungal assessment of NEEO were selected. The synthesis of evidence indicates that NEEO are effective in combating multidrug-resistant microorganisms and in the formation of their biofilms. Factors such as NE droplet size, chemical composition of essential oils, and the association of NE with antibiotics are discussed. Furthermore, NEEO showed satisfactory results in vitro and in vivo evaluations against resistant clinical isolates, making them promising for the development of new antimicrobial and antivirulence drugs.

Behavioral learning of dish rinsing and scrubbing based on interruptive direct teaching considering assistance rate

Robots are expected to manipulate objects in a safe and dexterous way. For example, washing dishes is a dexterous operation that involves scrubbing the dishes with a sponge and rinsing them with water. It is necessary to learn it safely without splashing water and without dropping the dishes. In this study, we propose a safe and dexterous manipulation system. The robot learns a dynamics model of the object by estimating the state of the object and the robot itself, the control input, and the amount of human assistance required (assistance rate) after the human corrects the initial trajectory of the robot's hands by interruptive direct teaching. By backpropagating the error between the estimated and the reference value using the acquired dynamics model, the robot can generate a control input that approaches the reference value, for example, so that human assistance is not required and the dish does not move excessively. This allows for adaptive rinsing and scrubbing of dishes with unknown shapes and properties. As a result, it is possible to generate safe actions that require less human assistance.

Ship autonomous collision avoidance decision from the perspective of navigation practice

Navigation safety is an eternal theme for ships, and it is an inherent strategic requirement for both transportation powers and maritime powers. With the development of intelligent ships, how to make intelligent decision-making results consistent with maritime practice is the key to navigation safety. This paper constructs a complex water area collision avoidance decision-making model that conforms to maritime practice, providing autonomous collision avoidance decision-making schemes for manned or unmanned ships. Firstly, based on the nonlinear velocity obstacle method, dynamic constraints such as ship maneuverability, target ship motion uncertainty, Convention on the International Regulations for Preventing Collisions at Sea (COLREGs), and good seamanship are fully considered, and a safe collision avoidance action plan is solved in the velocity domain; Then, by introducing an overshoot coefficient to improve the line of sight method, a feasible collision avoidance decision plan is derived by searching for a resumption navigation plan based on the safe avoidance action plan; Finally, collision avoidance decision plan is optimized to achieve autonomous collision avoidance decision-making for ships. To verify the effectiveness of the proposed model, a series of case studies, which focus on different encounter situations are designed and executed, the results indicate that the outcome of the model consistently aligns with real-world scenarios, which is in line with maritime practice.

The Effect of Stop Motion Video as An Examination Media on The Knowledge and Safe Behavior of PT. X Metal Casting Industry Workers in Klaten

A work accident is one of the risks that can occur in the metal casting industry. Workers' unsafe behavior is the primary cause of occupational accidents. The definition of unsafe behavior is any human behavior that can allow work accidents to occur to oneself or others. The preliminary study results found that 100% of workers had work accidents due to unsafe behavior, and out of 19 workers, 47.3% had insufficient knowledge about unsafe behavior. Counseling using video stop motion is one effort that can be used to overcome. This research aims to investigate the impact of counseling using video stop motion media on enhancing workers' knowledge and safety practices in the metal casting industry. This type of research is a quasi-experiment with a pre-test, post-test, and control group. The experimental group consisted of the workers in the finishing section, while the control group consisted of the workers in the induction section. The Wilcoxon test analysis reveals a significant difference in knowledge and behavior values between the pre-test and post-test, resulting from the use of video stop motion media. Media. The Mann-Whitney test showed a significant difference in the mean value of differences in knowledge and behavior in the pre-test and post-test between the experimental and control groups with a significance value of 0.000 (p <0.05). Video stop motion has the effect of increasing the value of knowledge and safe action among industry workers.

Effects of psilocybin on uncertain punishment learning

Psilocybin may provide a useful treatment for mood disorders including anxiety and depression but its mechanisms of action for these effects are not well understood. While recent preclinical work has begun to assess psilocybin’s role in affective behaviors through innate anxiety or fear conditioning, there is scant evidence for its role in conflict between reward and punishment. The current study was designed to determine the impact of psilocybin on the learning of reward-punishment conflict associations, as well as its effects after learning, in male and female rats. We utilized a chained schedule of reinforcement that involved execution of safe and risky reward-guided actions under uncertain punishment. Different patterns of behavioral suppression by psilocybin emerged during learning versus after learning of risky action-reward associations. Psilocybin increased behavioral suppression in female rats as punishment associations were learned. After learning, psilocybin decreased behavioral suppression in both sexes. Thus, psilocybin produces divergent effects on action suppression during approach-avoidance conflict depending on when the conflict is experienced. This observation may have implications for its therapeutic mechanism of action.

Analisis Tingkat Risiko Bahaya Area Kerja Laundrybox dengan Metode HIRARC

Laundry is a business sector that has the potential to cause the danger of work accidents. each potential danger has a different way of handling it. This research was conducted to find out the potential danger that could occur in a laundry business. HIRARC is a method used as a means of analyzing potential risks that may occur in work activities. From this research, it was found there are three activities or processes were identified as having potential dangers. These three activities consisted of spraying fragrances, using a clothes dryer, and inaccurate gas installation. Spraying fragrances has a high potential of danger with a risk rating value of 3 this potential danger can be prevented by using a facemask and glasses while working. The second activity is the use of a clothes dryer which has a low potential of danger with a risk rating value of 1, this activity can be prevented by implementing a safe action. The final work activity is a less careful gas installation with a moderate level of danger with a risk rating value of 2, more attention is needed when installing the regulator and using a good quality regulator. Keywords: HIRARC, laundry, risk

STPA-RL: Integrating Reinforcement Learning into STPA for Loss Scenario Exploration

Experience-based methods like reinforcement learning (RL) are often deemed less suitable for the safety field due to concerns about potential safety issues. To bridge this gap, we introduce STPA-RL, a methodology that integrates RL with System-Theoretic Process Analysis (STPA). STPA is a safety analysis technique that identifies causative factors leading to unsafe control actions and system hazards through loss scenarios. In the context of STPA-RL, we formalize the Markov Decision Process based on STPA analysis results to incorporate control algorithms into the system environment. The agent learns safe actions through reward-based learning, tracking potential hazard paths to validate system safety. Specifically, by analyzing various loss scenarios related to the Platform Screen Door, we assess the applicability of the proposed approach by evaluating hazard trajectory graphs and hazard frequencies in the system. This paper streamlines the RL process for loss scenario identification through STPA, contributing to self-guided loss scenarios and diverse system modeling. Additionally, it offers effective simulations for proactive development to enhance system safety and provide practical assistance in the safety field.

Safe Reinforcement Learning with Instantaneous Constraints: The Role of Aggressive Exploration

This paper studies safe Reinforcement Learning (safe RL) with linear function approximation and under hard instantaneous constraints where unsafe actions must be avoided at each step. Existing studies have considered safe RL with hard instantaneous constraints, but their approaches rely on several key assumptions: (i) the RL agent knows a safe action set for every state or knows a safe graph in which all the state-action-state triples are safe, and (ii) the constraint/cost functions are linear. In this paper, we consider safe RL with instantaneous hard constraints without assumption (i) and generalize (ii) to Reproducing Kernel Hilbert Space (RKHS). Our proposed algorithm, LSVI-AE, achieves O(√{d³H⁴K}) regret and O(H √{dK}) hard constraint violation when the cost function is linear and O(H?ₖ √{K}) hard constraint violation when the cost function belongs to RKHS. Here K is the learning horizon, H is the length of each episode, and ?ₖ is the information gain w.r.t the kernel used to approximate cost functions. Our results achieve the optimal dependency on the learning horizon K, matching the lower bound we provide in this paper and demonstrating the efficiency of LSVI-AE. Notably, the design of our approach encourages aggressive policy exploration, providing a unique perspective on safe RL with general cost functions and no prior knowledge of safe actions, which may be of independent interest.

Long-Term Safe Reinforcement Learning with Binary Feedback

Safety is an indispensable requirement for applying reinforcement learning (RL) to real problems. Although there has been a surge of safe RL algorithms proposed in recent years, most existing work typically 1) relies on receiving numeric safety feedback; 2) does not guarantee safety during the learning process; 3) limits the problem to a priori known, deterministic transition dynamics; and/or 4) assume the existence of a known safe policy for any states. Addressing the issues mentioned above, we thus propose Long-term Binary-feedback Safe RL (LoBiSaRL), a safe RL algorithm for constrained Markov decision processes (CMDPs) with binary safety feedback and an unknown, stochastic state transition function. LoBiSaRL optimizes a policy to maximize rewards while guaranteeing long-term safety that an agent executes only safe state-action pairs throughout each episode with high probability. Specifically, LoBiSaRL models the binary safety function via a generalized linear model (GLM) and conservatively takes only a safe action at every time step while inferring its effect on future safety under proper assumptions. Our theoretical results show that LoBiSaRL guarantees the long-term safety constraint, with high probability. Finally, our empirical results demonstrate that our algorithm is safer than existing methods without significantly compromising performance in terms of reward.

Learning Safe Action Models with Partial Observability

A common approach for solving planning problems is to model them in a formal language such as the Planning Domain Definition Language (PDDL), and then use an appropriate PDDL planner. Several algorithms for learning PDDL models from observations have been proposed but plans created with these learned models may not be sound. We propose two algorithms for learning PDDL models that are guaranteed to be safe to use even when given observations that include partially observable states. We analyze these algorithms theoretically, characterizing the sample complexity each algorithm requires to guarantee probabilistic completeness. We also show experimentally that our algorithms are often better than FAMA, a state-of-the-art PDDL learning algorithm.

Imitate the Good and Avoid the Bad: An Incremental Approach to Safe Reinforcement Learning

A popular framework for enforcing safe actions in Reinforcement Learning (RL) is Constrained RL, where trajectory based constraints on expected cost (or other cost measures) are employed to enforce safety and more importantly these constraints are enforced while maximizing expected reward. Most recent approaches for solving Constrained RL convert the trajectory based cost constraint into a surrogate problem that can be solved using minor modifications to RL methods. A key drawback with such approaches is an over or underestimation of the cost constraint at each state. Therefore, we provide an approach that does not modify the trajectory based cost constraint and instead imitates "good" trajectories and avoids "bad" trajectories generated from incrementally improving policies. We employ an oracle that utilizes a reward threshold (which is varied with learning) and the overall cost constraint to label trajectories as "good" or "bad". A key advantage of our approach is that we are able to work from any starting policy or set of trajectories and improve on it. In an exhaustive set of experiments, we demonstrate that our approach is able to outperform top benchmark approaches for solving Constrained RL problems, with respect to expected cost, CVaR cost, or even unknown cost constraints.

Autonomous Behavior Selection For Self-driving Cars Using Probabilistic Logic Factored Markov Decision Processes

ABSTRACT We propose probabilistic logic factored Markov decision processes (PL-fMDPs) as a behavior selection scheme for self-driving cars. Probabilistic logic combines logic programming with probability theory to achieve clear, rule-based knowledge descriptions of multivariate probability distributions, and a flexible mixture of deductive and probabilistic inferences. Factored Markov decision processes (fMDPs) are widely used to generate reward-optimal action policies for stochastic sequential decision problems. For evaluation, we developed a simulated self-driving car with reliable modules for behavior selection, perception, and control. The behavior selection module is composed of a two-level structure of four action policies obtained from PL-fMDPs. Three main tests were conducted focused on the selection of the appropriate actions in specific driving scenarios, and the overtaking of static obstacle vehicles and dynamic obstacle vehicles. We performed 520 repetitions of these tests. The self-driving car completed its task without collisions in 99.2% of the repetitions. Results show the suitability of the overall self-driving strategy and PL-fMDPs to construct safe action policies for self-driving cars.

Residual Physics and Post-Posed Shielding for Safe Deep Reinforcement Learning Method.

Deep reinforcement learning (DRL) has been researched for computer room air conditioning unit control problems in data centers (DCs). However, two main issues limit the deployment of DRL in actual systems. First, a large amount of data is needed. Next, as a mission-critical system, safe control needs to be guaranteed, and temperatures in DCs should be kept within a certain operating range. To mitigate these issues, this article proposes a novel control method RP-SDRL. First, Residual Physics, built using the first law of thermodynamics, is integrated with the DRL algorithm and a Prediction Model. Subsequently, a Correction Model adapted from gradient descent is combined with the Prediction Model as Post-Posed Shielding to enforce safe actions. The RP-SDRL method was validated using simulation. Noise is added to the states of the model to further test its performance under state uncertainty. Experimental results show that the combination of Residual Physics and DRL can significantly improve the initial policy, sample efficiency, and robustness. Residual Physics can also improve the sample efficiency and the accuracy of the prediction model. While DRL alone cannot avoid constraint violations, RP-SDRL can detect unsafe actions and significantly reduce violations. Compared to the baseline controller, about 13% of electricity usage can be saved.

Confirming the safety improvement and evacuation time reduction effects by the method for evacuating inundated areas via the shortest possible route

It is important to take evacuation behavior when a small river overflows or inland flooding occurs and there is a risk of subsequent flooding of a large river. To this end, even if roads are flooded, people should consider taking horizontal evacuation measures if possible and evacuating to places with a low risk of flooding or to evacuation shelters. Although local governments have published hazard maps for flood damage, these are mostly for large rivers and only provide information on the maximum depth of inundation. In addition, no evacuation routes have been described, and evacuees often choose the shortest route to evacuation shelters. However, if flooding significantly affects the shortest route to an evacuation shelter, then that is not the best route. Therefore, it is important to consider multiple evacuation routes that take flooding conditions into account. Thus, this study aims to validate whether method for evacuating inundated areas via the shortest possible route help improve evacuation safety and reduce evacuation time. The validation was conducted by means of a multiagent model evacuation simulation considering the time-series inundation depth of the inundation simulation results. After performing validation under multiple conditions, this study found that evacuating inundated areas via the shortest possible route not only increases evacuation safety but also significantly reduces the evacuation time required. The verification results show that method for evacuating inundated areas through the shortest route can be positioned as a safe and optimal evacuation action under occasionally changing circumstances when evacuation decisions are delayed.

Dual-enzyme decorated semiconducting polymer nanoagents for second near-infrared photoactivatable ferroptosis-immunotherapy.

Enzymes provide a class of potential options to treat cancer, while the precise regulation of enzyme activities for effective and safe therapeutic actions has been poorly reported. Dual-enzyme decorated semiconducting polymer nanoagents for second near-infrared (NIR-II) photoactivatable ferroptosis-immunotherapy are reported in this study. Such nanoagents (termed SPHGA) consist of hemoglobin (Hb)-based semiconducting polymer (SP@Hb), adenosine deaminase (ADA) and glucose oxidase (GOx) with loadings in a thermal-responsive nanoparticle shell. NIR-II photoactivation of SPHGA results in the generation of heat to trigger on-demand releases of two enzymes (ADA and GOx) via destroying the thermal-responsive nanoparticle shells. In the tumor microenvironment, GOx oxidizes glucose to form hydrogen peroxide (H2O2), which promotes the Fenton reaction of iron in SP@Hb, resulting in an enhanced ferroptosis effect and immunogenic cell death (ICD). In addition, ADA degrades high-level adenosine to reverse the immunosuppressive microenvironment, thus amplifying antitumor immune responses. Via NIR-II photoactivatable ferroptosis-immunotherapy, SPHGA shows an improved effect to absolutely remove bilateral tumors and effectively suppress tumor metastases in subcutaneous 4T1 breast cancer models. This study presents a dual-enzyme-based nanoagent with controllable therapeutic actions for effective and precise cancer therapy.