Deep Learning For Vision Tactile Sensing

Researchers at ETH Zurich have recently introduced a new deep learning-based strategy that could enable tactile sensing in robots without requiring large amounts of real-world data. In their experiments, researchers used a sensor they built with simple and low-cost components. This sensor is comprised of a standard camera placed below a soft material, which contains a random spread of tiny plastic particles. When a force is applied to its surface, the soft material deforms and causes the plastic particles to move. This motion is then captured by the sensor's camera and recorded. Researchers created models of the sensor's soft material and camera projection using state-of-the-art computational methods. They then used these models in simulations, to create a dataset of 13,448 synthetic images that is ideal for training tactile sensing algorithms. 

The fact that they were able to generate training data for their tactile sensing model in simulations is highly advantageous, as it prevented them from having to collect and annotate data in the real world. The researchers used the synthetic dataset they created to train a neural network architecture for vision-based tactile sensing applications and then evaluated its performance in a series of tests. The neural network achieved remarkable results, making accurate sensing predictions on real data, even if it was trained on simulations. In the future, the deep learning architecture could provide robots with an artificial sense of touch, potentially enhancing their grasping and manipulation skills. In addition, the synthetic dataset they compiled could be used to train other models for tactile sensing or may inspire the creation of new simulation-based datasets.

More information:

https://techxplore.com/news/2020-03-deep-learning-based-method-vision-based-tactile.html

AI Finds Disease-Related Genes

An artificial neural network can reveal patterns in huge amounts of gene expression data, and discover groups of disease-related genes. The scientists hope that the method can eventually be applied within precision medicine and individualised treatment. The researchers used AI to investigate whether it is possible to discover biological networks using deep learning, in which entities known as artificial neural networks are trained by experimental data. Since artificial neural networks are excellent at learning how to find patterns in enormous amounts of complex data, they are used in applications such as image recognition. However, this machine learning method has until now seldom been used in biological research. The scientists used a large database with information about the expression patterns of 20,000 genes in a large number of people. 

The information was unsorted, and the AI model was then trained to find patterns of gene expression. When they analysed their neural network, it turned out that the first hidden layer represented to a large extent interactions between various proteins. Deeper in the model, they found groups of different cell types. The scientists then investigated whether their model of gene expression could be used to determine which gene expression patterns are associated with disease and which is normal. They confirmed that the model finds relevant patterns that agree well with biological mechanisms in the body. Since the model has been trained using unclassified data, it is possible that the artificial neural network has found totally new patterns. The researchers plan now to investigate whether such, previously unknown patterns, are relevant from a biological perspective.

More information:

https://www.sciencedaily.com/releases/2020/02/200213132615.htm

Connecting Brain To Silicon-Based Technologies

Researchers at Stanford University have developed a new device for connecting the brain directly to silicon-based technologies. While brain-machine interface devices already exist and are used for prosthetics, disease treatment and brain research this latest device can record more data while being less intrusive than existing options. The device contains a bundle of microwires, with each wire less than half the width of the thinnest human hair. These thin wires can be gently inserted into the brain and connected on the outside directly to a silicon chip that records the electrical brain signals passing by each wire like making a movie of neural electrical activity. 

Current versions of the device include hundreds of microwires but future versions could contain thousands. The researchers tested their brain-machine interface on isolated retinal cells from rats and in the brains of living mice. In both cases, they successfully obtained meaningful signals across the array's hundreds of channels. Ongoing research will further determine how long the device can remain in the brain and what these signals can reveal. The team is especially interested in what the signals can tell them about learning. The researchers are also working on applications in prosthetics, particularly speech assistance.

More information:

https://techxplore.com/news/2020-03-device-silicon-power-brain-prosthetics.html

Wearable Biosensors For Personalized Health & Wellness

Bulky, buzzing and beeping hospital rooms demonstrate that monitoring a patient's health status is an invasive and uncomfortable process, at best, and a dangerous process, at worst. Penn State researchers want to change that and make biosensors that could make health monitoring less bulky, more accurate and much safer. The key would be making sensors that are so stretchable and flexible that they can easily integrate with the human body's complex, changing contours. If biosensors that are both energy efficient and stretchable can be achieved at scale, the researchers suggest that engineers can pursue a range of options for sensors that can be worn on the body, or even placed inside the body.

The payoff would be smarter, more effective and more personalized medical treatment and improved health decision-making without a lot of bulky, buzzing and beeping pieces of monitoring equipment. Some of the ideas that researchers at Penn State and around the world are investigating include stretchable textiles that can incorporate biosensors. Paper-based sensors could also potentially be used to create smart bandages that can monitor the status of wounds. Temporary tattoos could even incorporate biosensors for health monitoring. For example, a biosensor-enabled tattoo could provide diabetes patients with instant estimates of their glucose levels.

More information:

https://techxplore.com/news/2020-03-wearable-biosensors-pave-personalized-health.html

Shape-Changing, Free-Roaming Soft Robot

Stanford University researchers have developed a new kind of soft robot that, by borrowing features from traditional robotics, is safe while still retaining the ability to move and change shape. From that starting point, the researchers ended up with a human-scale soft robot that can change its shape, allowing it to grab and handle objects and roll in controllable directions. The simplest version of this squishy robot is an inflated tube that runs through three small machines that pinch it into a triangle shape. One machine holds the two ends of the tube together; the other two drive along the tube, changing the overall shape of the robot by moving its corners. The researchers call it an isoperimetric robot because, although the shape changes dramatically, the total length of the edges remains the same.

The isoperimetric robot is a descendent of three types of robots: soft robots, truss robots and collective robots. Soft robots are lightweight and compliant, truss robots have geometric forms that can change shape and collective robots are small robots that work together, making them particularly strong in the face of single-part failures. To make a more complex version of the robot, the researchers simply attach several triangles together. By coordinating the movements of the different motors, they can cause the robot to perform different behaviors, such as picking up a ball by engulfing it on three sides or altering the robot's center of mass to make it roll. For now, the researchers are experimenting with different shapes for their supple robot and considering plopping it in water to see if it can swim. They are also exploring even more new soft robot types, each with their own features and benefits.

More information:

https://techxplore.com/news/2020-03-stanford-shape-changing-free-roaming-soft-robot.html