New insight into non-isothermal crystallization of PVA-graphene composites

The melt crystallization of poly(vinyl alcohol) (PVA) and PVA composites has been a controversial subject due to inconclusive evidence and different opinions for its decomposition during crystallization. Using graphene as a model, the melt crystallization of PVA and PVA-graphene composites occurring during single-cycle and multiple-cycle non-isothermal annealing processes was systematically analyzed using different characterization techniques. The results obtained using single-cycle non-isothermal annealing indicated that the entire crystallization process took place through two main stages. The graphene in the PVA matrix regulates the nucleation and crystal growth manner of the PVA, yet resulting in retardation of the entire crystallization. The FTIR and Raman spectroscopic results particularly demonstrated that the annealing process not only improved the crystallinity but also led to clear decomposition in PVA and PVA-graphene composites, such as the elimination of hydroxyl groups and the production of C=C double bonds. The newly produced C=C double bonds were found to be responsible for the retardation of PVA macromolecule crystallization and the breaking of hydrogen bonds among the hydroxyl groups in the PVA chains. In addition, the morphological observation and multi-cycle non-isothermal crystallization further confirmed the existence of decomposition based on the surface damage as well as decreased crystallization enthalpy and crystallization peak temperature. Therefore, the non-isothermal crystallizations of the pure PVA and the PVA-graphene composites were in fact the combination of non-isothermal crystallization and non-isothermal degradation processes.

Dynamic virtual workspace for teleoperation of heterogeneous robots in non-deterministic environments

Robots are ever increasing in a variety of different workplaces providing an array of benefits such alternative solutions to traditional human labor. While developing fully autonomous robots is the ultimate goal in many robotic applications the reality is that there still exist many situationswere robots require some level of teleoperation in order to achieve assigned goals especially when deployed in non-deterministic environments. For instance teleoperation is commonly used in areas such as search and rescue, bomb disposal and exploration of inaccessible or harsh terrain. This is due to a range of factors such as the lack of ability for robots to quickly and reliably navigate unknown environments or provide high-level decision making especially intime critical tasks. To provide an adequate solution for such situations human-in-the-loop control is required. When developing human-in-the-loop control it is important to take advantage of the complimentary skill-sets that both humans and robots share. For example robots can performrapid calculations, provide accurate measurements through hardware such as sensors and store large amounts of data while humans provide experience, intuition, risk management and complex decision making capabilities. Shared autonomy is the concept of building robotic systems that take advantage of these complementary skills-sets to provide a robust an efficient robotic solution. While the requirement of human-in-the-loop control exists Human Machine Interaction (HMI) remains an important research topic especially the area of User Interface (UI) design.In order to provide operators with an effective teleoperation system it is important that the interface is intuitive and dynamic while also achieving a high level of immersion. Recent advancements in virtual and augmented reality hardware is giving rise to innovative HMI systems. Interactive hardware such as Microsoft Kinect, leap motion, Oculus Rift, Samsung Gear VR and even CAVE Automatic Virtual Environments [1] are providing vast improvements over traditional user interface designs such as the experimental web browser JanusVR [2]. This combined with the introduction of standardized robot frameworks such as ROS and Webots [3] that now support a large number of different robots provides an opportunity to develop a universal UI for teleoperation control to improve operator efficiency while reducing teleoperation training.This research introduces the concept of a dynamic virtual workspace for teleoperation of heterogeneous robots in non-deterministic environments that require human-in-the-loop control. The system first identifies the connected robots through the use kinematic information then determines its network capabilities such as latency and bandwidth. Given the robot type and network capabilities the system can then provide the operator with available teleoperation modes such as pick and place control or waypoint navigation while also allowing them to manipulate the virtual workspace layout to provide information from onboard camera’s or sensors.