This article presents a breakthrough in the area of Printed Electronics, by introducing a sintering technique that turns the nonconductive printed circuits into conductive and very stretchable circuits. On Nanoparticle Inks from silver, an increase in conductivity of 6 orders of magnitude and an increase in stretchability of over 25 times is reported.
This enhancement is achieved through a room temperature “sintering” process in which the liquid-phase EGaIn alloy binds the AgNP particles (~100 nm diameter) to form a continuous conductive trace. Ultra-thin and hydrographically transferrable electronics are produced by printing traces with a composition of AgNP-Ga-In on a 5μm thick temporary tattoo paper (TTP). The printed circuit is flexible enough to remain functional when deformed and can support strains above 80% with modest electromechanical coupling (gauge factor ~ 1). These mechanically robust thin-film circuits are well suited for transfer to highly curved and non-developable 3D surfaces as well as skin and other soft deformable substrates. In contrast to other stretchable tattoo-like electronics, the low-cost processing steps introduced here eliminate the need for cleanroom fabrication and instead requires only a commercial desktop printer. Most significantly, it allows for functionalities like “electronic tattoos” and 3D hydrographic transfer that have not been previously reported with EGaIn or EGaIn-based biphasic electronics.
Arthroscopy is a modality of orthopeadic surgery in which instruments and endoscopic camera (the arthroscope) are inserted into the articular cavity through small incisions (the surgical ports). Arthroscopy is highly beneficial for the patient and healthcare system because it reduces trauma, risk of infection and recovery time. However, clinical execution is difficult to accomplish because of indirect visualization and limited manoeuvrability inside the joint, with novices having to undergo a long training period and experts making mistakes of clinical consequences. This is a scenario where surgical assistive technologies can have strong impact in improving clinical outcome and disseminating the benefits of arthroscopy by increasing the number of adopters.
The ISR-UC created the first effective concept for accomplishing navigated arthroscopy. The solution combines real-time video processing for accurate 3D measurements on the anatomy, with augmented reality for overlaying meaningful guidance information in images. It is the first of the kind not requiring additional intra-operative sensing modalities, such as opto/magnetic tracking, that preclude the application in arthroscopy. Moreover, the improved usability, higher metric accuracy, and avoidance of additional capital equipment make video-based navigation also appealing for open orthopeadic surgery for which there are other competing systems.
Artificial perception plays a key role in the development of Intelligent Systems and Robotics. As a spin-off of computer vision and pattern recognition, there has been a focus on artificial multimodal perception for cognitive systems. The sensation-cognition-action loop needs to deal with uncertainty, and probabilistic approaches provide a robust solution, so novel solutions in Bayesian Computation have been pursued.
A key contribution in artificial multimodal perception for cognitive systems was on the actual way Bayesian computations are done. We are developing artificial perception algorithms, including integrated mult-isensory computational models. These include probabilistic approaches towards a new generation of algorithms to deal with uncertainty, ambiguities and conflicts inherent to the perceptual process that promote intelligent and adaptive decisions on actions in the physical world.
The main goals are:
- Integrated multisensory computational models and systems;
- Intelligent and adaptive cognitive decision and action/actuation processes;
- Computational models and devices to deal with uncertainty, ambiguities and conflicts inherent to the perceptual process.
We proposed the first successful theoretical extension of the standard Discriminative Correlation Filters, the Kernelized Correlation Filter (KCF), which achieved remarkable tracking performance, but still preserved high speed.
We explored the structure of circulant matrices for the purpose of detection and tracking, detailing their relationship to cyclic shifts and the Discrete Fourier Transform (DFT). The main idea put forward is that cyclic shifts provide an accurate model for the sampling process inherent in recognition algorithms, and that these algorithms can be greatly accelerated by leveraging properties of cyclic shifts. We conducted a systematic study of several algorithms and settings, showing that they exhibit circulant structure under this model, and proposing novel solutions that are orders of magnitude faster than the state-of-the-art. We demonstrated, for the first time, the connection between ridge regression with cyclically shifted samples and classical correlation filters. Conceptually, this achievement is considered the major contribution to the success of correlation-based trackers. We also proposed, for the first time, the use of multiple channels in correlation filters, which has become very popular in visual tracking. Minimalistic trackers based on Kernelized Correlation Filters (KCF) and Dual Correlation Filters (DCF) was proposed, which run at hundreds of frames-per-second, and yet perform better than more complex trackers on challenging benchmarks.
Cardiac interventions have to deal with several difficulties, where heart motion is a major one. This work addresses the heart motion compensation problem under contact, relying on robot force control techniques. A double active observer (AOB) architecture is proposed to tackle precise force control in the presence of heart motion. One AOB controls the desired interaction force, and the other one is responsible for compensating heart motion autonomously. 3-DOF force tracking on top of a beating heart is presented, showing that with the proposed control architecture it is possible to have force controlled interventions with RMS errors below 0.56 [N] (based on raw data). The 4-DOF Whole Arm Manipulator (WAM) from Barrett Technologies is the medical arm, while another 3-DOF robot carrying an ex-vivo heart generates heart motion based on recorded biosignals.