A research was conducted by Schuster et al., (2009) for addressing the rising needs, sophistication, and complexity of robots, stakeholders in the automation and robotics industry are working for establishing new standards of international safety through the ISO or International Organization for Standardization for robot systems and robot integration. The initial standard, ISO 10218-1, specifies requirements while providing guidance for safety assurance in construction and design of the robot, not the overall system of the robot. ISO 10218-2 was expected to be established in 2011 and it is covering installation and integration of cell or robot system, thus offering a comprehensive set of guidelines for safety. Those issues of robotic safety and others are seemingly addressed in the current standard of safety, ANSI/RIA R15.06. However, those guidelines were selected in 1999 and they don’t cover innovations which have been developed since then (Schuster et al., 2009).

            In accordance with the authors, these advancements in integration technologies and capabilities offer various benefits. Other than helping in increasing safety and worker productivity, the promise of smaller footprint of robot-system is held by them than traditional technology, which is based on stops of mechanical safety, or external controls or sensors. PLCs or programmable logic controllers which related to safety play a significant role in cells of robotic work. Input data is collected from sensors from them about person status versus a robot within different spaces, along with inputs from different safety devices like interlock switches, position sensors, pendants, and e-stops. Output of PLC helps in controlling the circuit of robot power and robot servos, along with present motors, pneumatic and hydraulic devices. In addition, safety technology that connects robots directly to the safety bus has the potential of providing more granular information through interfaces between human-machine. Overall, information provided by advanced systems of safety contributes to the initiatives of continuous improvement by evaluating the failures and faults of robot systems on a historical and statistical basis. For instance, if managers recognize that a specific component of safety fails more often than many others, the issue can be resolved for saving the money and time of maintenance (Schuster et al., 2009).

            In accordance with a study of Woodman et al. (2012), there has been a significant effort for addressing the safety issues related to pHRI or physical robot-human interaction. But several challenges are still remaining. For different personal robots and the ones which are expected to work in unstructured environments, the safety issue is compounded. It has been argued by authors that traditional design techniques are unable to identify the complexities which are related to dynamic environments. An overview of our control system was prevented by authors along with its methodology of implementation. It will subsequently serve as an enforcer of high-level safety by seemingly governing the robotic actions and limiting the control layer from carrying out unsafe operations. For demonstrating the design effectiveness, different experiments have been carried out with the use of MobileRobots PeopleBot. Lastly, outcomes are presented with the demonstration of how failures injected into a specific controller can be handled and identified by the system of safety protection (Woodman et al., 2012).

            In accordance with the study carried out by Salau (2015), UAVs or Unmanned Aerial Vehicles utilization for different indoor operations is increasing in demand with its application in the recovery of disaster, inspection in the industry of manufacturing, and in the system of healthcare. That is why it is important to establish a scheme of obstacle avoidance for different UAVs flying under a specific altitude. Multi-obstacle avoidance has been focused upon by this work for UAVs quadcopter type. The strategy of SA or Simple Navigation Algorithm utilized the potential of artificial navigation for generating new point way for the robot. MATLAB was utilized for verifying and stimulating the approach (Salau, 2015).

Human Locomotion Study of Interaction between Humans and Robots

            A research was conducted by Arechavaleta et al. (2006) which proposes a differential system which describes precisely the design of human walking’s locomotor trajectories on the ground level, without obstacles. The approach of the author emphasizes the close relationship between the kinematic model of the mobile robot and locomotor path shape in movements which are goal-directed.  It is indicated by this observation that some limitations act on bodies of humans which limit the way how locomotor trajectories are generated by humans (Arechavaleta et al., 2006).

            A differential system was proposed by Arechavaleta et al. (2006) which respects all the nonholonomic limitations. This model is validated by authors by comparing different stimulated trajectories with the recorded trajectories which are developed during the locomotion which is goal-oriented in humans. All the subjects had to begin from pre-determined directions and positions for crossing over a specific porch (orientation and position of the porch were the two factors of manipulation). On a database of trajectories reaching 1,560 were recorded from 7 subjects. A promising route is opened by it for understanding human locomotion through tools of differential geometry experienced in robotics of mobile (Arechavaleta et al., 2006).

              A research was conducted by Lin and Pandy (2017) for performing three-dimensional simulations of the human locomotion by driving a model of neuromusculoskeletal toward in different vivo measurements of ground reaction and body-segmental forces of kinematics. Interaction between ground and foot was simulated with the use of 6 contact spheres under each and every foot. The issue of dynamic optimization was to identify the group of muscle excitations required for reproducing ground reaction forces’ 3D measurements while minimizing the squared muscle activations. 2.7 ± 1.0 h was the time along with 2.2 ± 1.6 h which was taken by direct collocation of CPU time for solving the issues of optimization for running and walking. Forces between foot-ground and computed kinematics were in a well agreement with the experimental data while the evaluated patterns of muscle excitation were consistent with the activity of EMG (Lin and Pandy, 2017).

            In accordance with the study of Haghpanah et al (2007), CPG or central pattern generators in the spinal cord are considered to be accountable for developing the patterns of the rhythmic motor during the rhythmic activities. A CPG model of four neurons presented by Matsuoka was utilized with mutual inhibition for generating muscle synergies’ activation patterns, utilizing the information of foot contact and flexion angle of him from inputs of sensory afferents (Haghpanah et al, 2007).

                Model parameters were tuned by Haghpanah et al (2007) using an individual gait trial’s experimental data, which resulted in a good accuracy of fitting (RMSEs among 0.1399 and 0.0491) between the activations of experimental synergy and simulation. Then the performance of the model was assessed by the comparison of its predictions for the patterns of activation of individual muscles of the leg during locomotion with relative data of EMG. It was indicated by results that characteristic features of complex patterns of activation of muscles were reproduced well by the model for different subjects and gait trials. Generally, the muscle synergy and CPG-based model was quite promising considering its simple yet extensive potential for a reliable neuromuscular control such as resolving redundancies, fast and distributed control, and locomotion modulation by simple signals of control (Haghpanah et al, 2007).