Search results

The purpose of this systematic review is to critically analyze pharmaceutical industry case studies on the implementation of Lean 4.0 methodologies to synthesize key lessons, benefits and best practices. The goal is to inform decisions and guide investments in related technologies for enhancing quality, compliance, efficiency and responsiveness across production and supply chain processes.

The article utilized a systematic literature review (SLR) methodology following five phases: formulating research questions, locating relevant articles, selecting and evaluating articles, analyzing and synthesizing findings and reporting results. The SLR aimed to critically analyze pharmaceutical industry case studies on Lean 4.0 implementation to synthesize key lessons, benefits and best practices.

Key findings reveal recurrent efficiency gains, obstacles around legacy system integration and data governance as well as necessary operator training investments alongside technological upgrades. On average, quality assurance reliability improved by over 50%, while inventory waste declined by 57% based on quantified metrics across documented initiatives synthesizing robotics, sensors and analytics.

As a comprehensive literature review, findings depend on available documented implementations within the search period rather than direct case evaluations. Reporting bias may also skew toward more successful accounts.

Synthesized implementation patterns, performance outcomes and concealed pitfalls provide pharmaceutical leaders with an evidence-based reference guide aiding adoption strategy development, resource planning and workforce transitioning crucial for Lean 4.0 assimilation.

This systematic assessment of pharmaceutical Lean 4.0 adoption offers an unprecedented perspective into the real-world issues, dependencies and modifications necessary for successful integration, absent from conceptual projections or isolated case studies alone until now.

Maintaining the safety of the human is a major concern in factories where humans co-exist with robots and other physical tools. Typically, the area around the robots is monitored using lasers. However, lasers cannot distinguish between human and non-human objects in the robot’s path. Stopping or slowing down the robot when non-human objects approach is unproductive. This research contribution addresses that inefficiency by showing how computer-vision techniques can be used instead of lasers which improve up-time of the robot.

A computer-vision safety system is presented. Image segmentation, 3D point clouds, face recognition, hand gesture recognition, speed and trajectory tracking and a digital twin are used. Using speed and separation, the robot’s speed is controlled based on the nearest location of humans accurate to their body shape. The computer-vision safety system is compared to a traditional laser measure. The system is evaluated in a controlled test, and in the field.

Computer-vision and lasers are shown to be equivalent by a measure of relationship and measure of agreement. R² is given as 0.999983. The two methods are systematically producing similar results, as the bias is close to zero, at 0.060 mm. Using Bland–Altman analysis, 95% of the differences lie within the limits of maximum acceptable differences.

In this paper an original model for future computer-vision safety systems is described which is equivalent to existing laser systems, identifies and adapts to particular humans and reduces the need to slow and stop systems thereby improving efficiency. The implication is that computer-vision can be used to substitute lasers and permit adaptive robotic control in human–robot collaboration systems.

Limited by the types of sensors, the state information available for musculoskeletal robots with highly redundant, nonlinear muscles is often incomplete, which makes the control face a bottleneck problem. The aim of this paper is to design a method to improve the motion performance of musculoskeletal robots in partially observable scenarios, and to leverage the ontology knowledge to enhance the algorithm’s adaptability to musculoskeletal robots that have undergone changes.

A memory and attention-based reinforcement learning method is proposed for musculoskeletal robots with prior knowledge of muscle synergies. First, to deal with partially observed states available to musculoskeletal robots, a memory and attention-based network architecture is proposed for inferring more sufficient and intrinsic states. Second, inspired by muscle synergy hypothesis in neuroscience, prior knowledge of a musculoskeletal robot’s muscle synergies is embedded in network structure and reward shaping.

Based on systematic validation, it is found that the proposed method demonstrates superiority over the traditional twin delayed deep deterministic policy gradients (TD3) algorithm. A musculoskeletal robot with highly redundant, nonlinear muscles is adopted to implement goal-directed tasks. In the case of 21-dimensional states, the learning efficiency and accuracy are significantly improved compared with the traditional TD3 algorithm; in the case of 13-dimensional states without velocities and information from the end effector, the traditional TD3 is unable to complete the reaching tasks, while the proposed method breaks through this bottleneck problem.

In this paper, a novel memory and attention-based reinforcement learning method with prior knowledge of muscle synergies is proposed for musculoskeletal robots to deal with partially observable scenarios. Compared with the existing methods, the proposed method effectively improves the performance. Furthermore, this paper promotes the fusion of neuroscience and robotics.