A new hand training system has been designed and built to help clinicians administrate hand therapy to stroke patients. It uses pneumatics to actuate the fingers from the dorsal side of the hand in order to provide training for activities of daily living. Before the device is tested in a clinical trial, it’s safety, comfort and reliability needed to be established via trials on 30 healthy individuals. A comfort survey that was issued to the users indicated that they found the device comfortable regardless of age, gender, weight and hand length. The sensor data that was gathered during testing indicated that the readings were reliable and the device had minimal impact on the subjects' normal range of motion. A bio-mechanical model, validated through experimentation, was also created to estimate joint angles of the index finger during the trials to ensure that the device put the joints of the finger in bio-mechanically safe angles. / Natural Sciences and Engineering Research Council of Canada

Recent studies have demonstrated that images constructed from lidar reflectance information exhibit superior robustness to lighting changes. However, due to the scanning nature of the lidar and assumptions made in previous implementations, data acquired during continuous vehicle motion suffer from geometric motion distortion and can subsequently result in poor metric visual odometry (VO) estimates, even over short distances (e.g., 5-10 m). The first part of this thesis revisits the measurement timing assumption made in previous systems, and proposes a frame-to-frame VO estimation framework based on a pose-interpolation scheme that explicitly accounts for the exact acquisition time of each intrinsic, geometric feature measurement. The second part of this thesis investigates a novel method of lidar calibration that can be applied without consideration of the internal structure of the sensor. Both methods are validated using experimental data collected from a planetary analogue environment with a real scanning laser rangefinder.

The rationale behind this research originates from the lack of public health care in South Africa. There is an escalation in the number of stroke victims which is a consequence of the increase in hypertension in this urbanising society. This increase results in a growing need for physiotherapists and occupational therapists in this country which is further hindered by the division between urban and rural areas. The exoskeleton device has been formulated to encapsulate methodologies that enable the anthropomorphic integration between a biological and mechatronic limb. The physiotherapeutic mechanism was designed to be portable and adjustable, without limiting the spherical motion and workspace of the human arm. The exoskeleton was portable in the sense that it could be transported geographically and is a complete device allowing for motion in the shoulder, elbow, wrist and hand joints. The avoidance of singularities in the workspace required the implementation of non-orthogonal joints which produces extensive forward kinematics. Traditional geometric or analytical derivations of the inverse kinematics are complicated by the nonorthogonal layout. This hindrance was resolved iteratively via the Damped Least Squares method. The electronic and computer system allowed for professional personnel, such as an occupational therapist or a physiotherapist, to either change an individual joint or a combination of joints angles. A ramp PI controller was established to provide a smooth response in order to simulate the passive therapy motion. / Thesis (M.Sc.Eng)-University of KwaZulu-Natal, Durban, 2011.

Recent studies have demonstrated that images constructed from lidar reflectance information exhibit superior robustness to lighting changes. However, due to the scanning nature of the lidar and assumptions made in previous implementations, data acquired during continuous vehicle motion suffer from geometric motion distortion and can subsequently result in poor metric visual odometry (VO) estimates, even over short distances (e.g., 5-10 m). The first part of this thesis revisits the measurement timing assumption made in previous systems, and proposes a frame-to-frame VO estimation framework based on a pose-interpolation scheme that explicitly accounts for the exact acquisition time of each intrinsic, geometric feature measurement. The second part of this thesis investigates a novel method of lidar calibration that can be applied without consideration of the internal structure of the sensor. Both methods are validated using experimental data collected from a planetary analogue environment with a real scanning laser rangefinder.

Spending on robot surgery is expected to increase by $17 billion in the next 6 years. This new surgical treatment has challenged hospitals with higher costs and varying performance. Healthcare executives struggle balancing the adoption of medical innovations with managing healthcare costs. This dilemma can be further complicated by industry structures relative to capital-intensive medical innovations. This research explores the interaction between industry structure and customer value. Specifically, how can hospitals apply an understanding of supplier industry structure and customer value to improve the value of a robotic surgery program (RSP)? This industry study represents an exhaustive longitudinal review of over 15 years of public data relative to robotic surgery, across three distinct time periods. Within the research, industry structure is evaluated using Porter’s 5-forces model. A framework based upon contributions from Grönroos as well as Menon, Homburg, and Beutin is introduced to assess customer value based upon clinical, financial and strategic (CFS) value. The implications of periodic industry structure on customer value were examined to identify opportunities for hospital executives to increase RSP customer value. There were several empirical and theoretical findings from this research. First, in the face of increasing industry structure the identification of favorable forces may create opportunities to increase RSP value. Secondarily, exploring customer value through the lens of core, add-on, relational and transactional benefits in the sub-context of CFS value aids in the identification of market power influences on customer value. The implications of the absence of high levels of relational and transactional benefits without high levels of core and add-on benefits may influence avenues of pursuit in improving RSP value overall. The research also suggests that clinical and strategic value was present despite varying degrees of industry structure. Finally, this study represents an empirical joint analysis of industry structure and customer value in robotic surgery. Some proponents may find the introduction of an integrative model for measuring customer value in robotic surgery, applicable to other capital-intensive medical innovations or disruptive technologies at large.

This thesis presents the development of a novel robotically effected plastic foam sculpting system for rapid prototyping and manufacturing purposes. The developed system is capable of rapidly sculpting physical objects out of expanded and extruded polystyrene using an electrically heated Nichrome sculpting tool. An overview of current conventional rapid prototyping systems indicated that the main disadvantages lie in the limited size of objects which can be built, the relatively long time involved to produce one part and the high cost of the systems and materials. An extensive literature and technology review was conducted on work which was similar to the novel system presented in this thesis. The literature provided many good ideas which could be applied. Two sections of experimental work were conducted. The first was aimed at simply proving the concept of robotically effected sculpting of plastic foams. A crude procedure was developed which proved to be rather tedious and manual, especially in terms of generating the tool paths. Qualitative observations of the cut surfaces were used to change the testing parameters to explore their effects and discover which parameters produced accurate and smooth sculpted surfaces. 12 tests were documented and proved that the sculpting of satisfactory surfaces was achievable. The second section of experimental work involved developing the aforementioned crude procedure to make it more automated, especially in terms of the tool path generation and optimisation step. An innovative five step procedure was developed which if followed can produce accurately sculpted artefacts using CAD models of the artefacts as the primary input. Two artefacts were successfully sculpted using the developed procedure. The first was a simple lofted surface; the CAD model of which was created in SolidWorks. The second artefact was a patient customised medical radiation therapy head and neck support; the CAD model of which was created by scanning the back of the author's head and neck with a 3D scanner. The sculpted support fitted the author perfectly. The implementation of the procedure in the two tests highlighted several points including the speed in which the whole process can be carried out. The time taken from the scanning of the authors head and neck with the 3D scanner through to the physical sculpted artefact, was a mere 80 minutes; of which only 13 minutes was consumed in the actual setup and sculpting step! This is extremely quick when compared to conventional rapid prototyping systems and CNC milling. Several areas of future work were outlined and included, tool and fixture design, automation and integration of the system procedure, tool pathing strategy for foam cutting and robot control system issues. The work presented in this thesis provides an excellent foundation for future development of the robotic foam sculpting system.

The purpose of this project is the creation of an adaptive Function Block control system, and the implementation of Artificial Intelligence integrated within the Function Block control system, using IEC 61499 standard to control an ABB 6-axis virtual robot, simulated in the software RobotStudio. To develop these objectives, we studied a lot of necessary concepts and how to use three different softwares. To learn how to use the softwares, some tests were carried out. RobotStudio is a program developed by ABB Robotics Company where an ABB robot and a station are simulated. There, we designed and created a virtual assembly cell with the virtual IRB140 robot and the necessary pieces to simulate the system. To control the robot and the direct access to the different tools of RobotStudio, it is necessary to use an application programming interface (API) developed by ABB Robotics Company. C sharp (C#) language is used to program using the API, but this language is not supported by the Function Block programming software nxtStudio. Because of this, we used VisualStudio software. In this software, we use the API libraries to start and stop the robot and load a RAPID file in the controller. In a RAPID file the instructions that the robot must follow are written. So, we had to learn about how to program in C# language and how to use VisualStudio software. Also, to learn about IEC 61499 standard it was necessary to read some books. This standard determines how an application should be programmed through function blocks. A function block is a unit of program with a certain functionality which contains data and variables that can be manipulated in the same function block by several algorithms. To program in this standard we learnt how to use nxtStudio software, consuming a lot of time because the program is quite complex and it is not much used in the industrial world yet. Some tests were performed to learn different programing skills in this standard, such as how to use UDP communication protocol and how to program interfaces. Learning UDP communication was really useful because it is necessary for communication between nxtStudio and other programs, and also learning how to use interfaces to let the user access the program. Once we had learnt about how to use and program the different softwares and languages, we began to program the project. Then, we had some troubles with nxtStudio because strings longer than fourteen characters cannot be used here. So, a motion alarm was developed in VisualStudio program. And another important limitation of nxtStudio is that C++ language cannot be used. Therefore, the creation of an Artificial Intelligence system was not possible. So, we created a Function Block control system. This system is a logistical system realised through loops, conditions and counters. All this makes the robot more adaptive. As the AI could not be carried out because of the different limitations, we theoretically designed the AI system. It will be possible to implement the AI when the limitations and the problems are solved.

No description available.

We describe our overall approach to building robot vision systems, and the conceptual systems architecture as a network of agents, which run in parallel, and cooperate to achieve the system’s goals. We present the current state of the 3D Feature-Based Tracker, a robot vision system for tracking and segmenting the 3D motion of objects using image input from a calibrated stereo pair of video cameras. The system runs in a multi-level cycle of prediction and verification or correction. The currently modelled 3D positions and velocities of the feature points are extrapolated a short time into the future to yield predictions of 3D position. These 3D predictions are projected into the two stereo views, and are used to guide a fast and highly focused visual search for the feature points. The image positions at which the features are re-acquired are back-projected in 3D space in order to update the 3D positions and velocities. At a higher level, features are dynamically grouped into clusters with common 3D motion. Predictions from the cluster level can be fed to the lower level to correct errors in the point-wise tracking.

Robotic swarms are multi-robot systems whose global behavior emerges from local interactions between individual robots and spatially proximal neighboring robots. Each robot can be programmed with several local control laws that can be activated depending on an operator’s choice of global swarm behavior (e.g. flocking, aggregation, formation control, area coverage). In contrast to other multi-robot systems, robotic swarms are inherently scalable since they are robust to addition and removal of members with minimal system reconfiguration. This makes them ideal for applications such as search and rescue, environmental exploration and surveillance. Practical missions often require a combination of swarm behaviors and may have dynamically changing mission goals. However, a robotic swarm is a complex distributed dynamical system, so its state evolution depends on the timing as well as sequence of the supervisory inputs. Thus, it is difficult to predict the effects of an input on the state evolution of the swarm. More specifically, after becoming aware of a change in mission goals, it is unclear at what time a supervisory operator must convey this information to the swarm or which combination of behaviors to use to accomplish the new goals. The main challenges we address in this thesis are characterizing the effects of input timing on swarm performance and using this theory to inform automated composition of swarm behaviors to accomplish updated mission goals. We begin by formalizing the notion of Neglect Benevolence — the idea that delaying the application of an input can sometimes be beneficial to overall swarm performance — and using the developed theory to demonstrate experimentally that humans can learn to approximate optimal input timing. In an adversarial setting, we also demonstrate that by altering only the timing of consensus updates for a subset of the swarm, we can influence the agreement point of the entire swarm. Given a library of swarm behaviors, automated behavior composition consists of identifying a behavior schedule that must specify (1) the appropriate sequence of behaviors and (2) the corresponding duration of execution for each behavior. Applying our notion of Neglect Benevolence, it is clear these two parts are intricately interdependent. By first assuming the durations are known, we present an algorithm to identify the optimal behavior sequence to achieve a desired swarm mission goal when our library contains general swarm behaviors. By restricting our library to consensus-based swarm behaviors, we then relax the assumption on known durations and present an algorithm to simultaneously find the sequence and durations of swarm behaviors to time-optimally accomplish multiple unordered goals.