We present a bilateral teleoperation method capable of controlling magnetic microrobots. Our method is built on the MagnetoSuture, a quad-coil magnetic actuation system which controls magnetic fields to perform tetherless tissue suturing with sintered Neodymium Iron Boron (NdFeB) suture needles. While under visually rich conditions, we found that haptic feedback preference varies greatly, many magnetic robot procedures are conduced under slow, highly corrupted or otherwise severely limited visual feedback, such as MRI, CT Scans, and fluoroscopy. Here we discuss the design and architecture of this teleportation strategy, including detailed technical analysis of the FEM-based control method, Tikhonov regularization for improving performance under ill-posed conditions, EMI-resistant online force measurement, and fluoroscopy simulations for user studies. We also evaluate the efficacy of this control method in a mock aneurysm coiling task and discuss measures indicating both task performance and neurological performance.

Current commercially available robotic minimally invasive surgery (RMIS) platforms provide no haptic feedback of tool interactions with the surgical environment. As a consequence, novice robotic surgeons must rely exclusively on visual feedback to sense their physical interactions with the surgical environment. This technical limitation can make it challenging and time-consuming to train novice surgeons to proficiency in RMIS. Extensive prior research has demonstrated that incorporating haptic feedback is effective at improving surgical training task performance. However, few studies have investigated the utility of providing feedback of multiple modalities of haptic feedback simultaneously (multi-modality haptic feedback) in this context, and these studies have presented mixed results regarding its efficacy. Furthermore, the inability to generalize and compare these mixed results has limited our ability to understand why they can vary significantly between studies. Therefore, we have developed a generalized, modular multi-modality haptic feedback and data acquisition framework leveraging the real-time data acquisition and streaming capabilities of the Robot Operating System (ROS). In our preliminary study using this system, participants complete a peg transfer task using a da Vinci robot while receiving haptic feedback of applied forces, contact accelerations, or both via custom wrist-worn haptic devices. Results highlight the capability of our system in running systematic comparisons between various single and dual-modality haptic feedback approaches.

The Hapkit is an open-source kinesthetic device used by educators to teach fundamentals of haptic interaction. Despite its pedagogical benefits, it cannot showcase ungrounded interactions. Here, we introduce an ungrounded Hapkit variant that generates moments at the wrist as the user interacts with a virtual environment. The device utilizes existing Hapkit hardware with 3D printed modifications to enable ungrounded kinesthetic feedback. We developed Arduino and Processing code that simulates ungrounded impacts, such as hitting a tennis ball, to demonstrate functionality. This device can serve as a valuable pedagogical tool in conjunction with the well-established Hapkit.