Biomechatronic Systems

Human systems, and the individual cells that comprise them, are complex. The development of improved engineering methods to track and analyze them in an automated fashion has crucial applications from cancer screening to early diagnosis of a chronic condition or detection of an acute event, e.g., stroke, falling in older adults or during rehabilitation from a surgical procedure. Our current research efforts are specifically focused on the development of enabling devices, systems and techniques for automated tracking and analysis of moving human cells (e.g., circulating tumor and stem cells in cancer patients’ blood) and systems (e.g., human bodies and subjects, single or multiple), which aligns with BSL’s strategic vision of “Discovering the Meaning of Motion” and having an international reputation in this area. BSL's ongoing research can be divided into the below four main areas. The aim of this research is the development of a novel cell sorting, manipulation and analysis system dubbed “RoboSCell” [J17]. Compared to existing cell sorting methods that are targeted for bulk-cell or group-cell sorting, the proposed technology provides a novel single-cell sorting method that allows for automatic identification, isolation and arraying of individual cells, with applications towards monitoring infection events or pharmacological actions, stem cell and cancer cell analysis, all done at the single-cell level. The proposed immunomagnetic single cell arraying technique [J14] of RoboSCell can acquire time-based measurements (i.e. dynamic analysis) on the same cells, allowing for population-based comparisons, as we demonstrated in [J17]. Additionally, single cell electroporation (SCE) [J15], which is one of key capabilities developed for RoboSCell, opens up new opportunities for manipulating the genetic, metabolic, and synthetic contents of targeted individual cells for further processing and analysis. SCE is an emerging and powerful technique for localized gene transfection and microinjection of exogenous molecules into single cells. However, an absence of fully automated SCE of thin adherent cells reduces the generality of the technique. In order to address this issue, [J29] demonstrates a fully automated method for detecting microcapillary tip-membrane contact using statistical process control. In addition, [J27] presents a novel machine vision-based automated localization of SCE sites. Finally, [J34] successfully demonstrates a fully automated SCE process, delivering fluorescing dyes and reporter genes into micron-scale cellular features of adherent NIH/3T3 cells. The SCE technology is still in its infancy and we have achieved a significant new milestone in unleashing its true power by overcoming some of its major technical challenges and unknown variables involved in the execution. The aim of this research is the development of portable and wearable ambulatory monitoring systems using distributed and wireless miniature inertial measurement units (IMUs). The potential impact of such an ambulatory motion tracking system is significant for the practice of physical medicine and rehabilitation. For example, it opens the possibility of continuous monitoring for early diagnosis of a chronic condition, an acute event such as stroke or falling in older adults or during rehabilitation from a surgical procedure. We have proposed novel quaternion-based sensor fusion algorithms [J12-13] that accurately compute 3D orientation of body segments using IMUs. [J7] presents a novel application of the proposed system for real-time measurement of the complex 3D human spinal motion, for non-invasive diagnosis of spine injury/pain in clinics. [J22] is a subsequent study that investigates spinal motion pattern differences during various human gaits in real-life, which was followed by a clinical investigation of the proposed method’s effectiveness in assessing patients with low back problems in [J23]. [J21] is a novel real-time gait event detection method using shank-attached gyroscopes. [J26] presents comprehensive dynamic models of the 3D human posture and gait to accurately simulate their patterns, using IMUs or optical markers. Finally, [J33] presents a novel method to predict dynamic feet-ground interactions for analysis of the state of balance in bipedal (human or humanoid) walking, by redefining the zero moment point (ZMP) criterion. This research is aimed at the development of an automotive magnetorheological brake (MRB) system. The research issues being addressed include: MR fluid characterization and electromagnet design; magnetic circuit design and optimization, prototyping and manufacturing; and intelligent controller design and implementation. Two papers, [J6] and [J8] that present an MRB-based ABS controller and multidisciplinary design optimization (MDO) of the MRB, respectively, have been published. [J9] is a subsequent work on experimental prototyping and evaluation of the MRB. Both [J6] and [J9] have gathered significant interest in the literature (Top 10 most cited articles in Mechatronics between 2006 and 2011). As the next stage of this research, we are focusing on the performance improvements (e.g., braking torque and reliability) of the MRB (e.g., [J25]) and the development of a hybrid system that integrates an eddy current brake (ECB) [J32] or switched reluctance motor. This research addresses the problem of active control of flexible structures using a large array of sensors/actuators and distributed control techniques that allow unprecedented capabilities for controlling such vibration-prone structures. However, this also leads to a large number of inputs and outputs in the control system that is challenging for standard control techniques to handle. Therefore, our research is aimed at finding solutions to such challenges via recent advances in distributed control theories. Two specific and novel applications are currently being studied: (i) distributed control of large segmented primary mirrors of next-generation ground-based telescopes [J20, J28] and (ii) distributed control of very flexible manipulators [J11, J16] for underwater or space applications. Previously, a so-called "smart" robotic gripper with actuated contact fingers to grasp flexible payloads (e.g., automotive sheet metal parts) was developed by employing distributed sensing and actuation [J4-5].