AI-driven robotic chemist for autonomous synthesis of organic molecules.

The automation of organic compound synthesis is pivotal for expediting the development of such compounds. In addition, enhancing development efficiency can be achieved by incorporating autonomous functions alongside automation. To achieve this, we developed an autonomous synthesis robot that harnesses the power of artificial intelligence (AI) and robotic technology to establish optimal synthetic recipes. Given a target molecule, our AI initially plans synthetic pathways and defines reaction conditions. It then iteratively refines these plans using feedback from the experimental robot, gradually optimizing the recipe. The system performance was validated by successfully determining synthetic recipes for three organic compounds, yielding that conversion rates that outperform existing references. Notably, this autonomous system is designed around batch reactors, making it accessible and valuable to chemists in standard laboratory settings, thereby streamlining research endeavors.

Real-Time Automated Solubility Screening Method Using Deep Neural Networks with Handcrafted Features.

Solubility measurements are essential in various research and industrial fields. With the automation of processes, the importance of automatic and real-time solubility measurements has increased. Although end-to-end learning methods are commonly used for classification tasks, the use of handcrafted features is still important for specific tasks with the limited labeled images of solutions used in industrial settings. In this study, we propose a method that uses computer vision algorithms to extract nine handcrafted features from images and train a DNN-based classifier to automatically classify solutions based on their dissolution states. To validate the proposed method, a dataset was constructed using various solution images ranging from undissolved solutes in the form of fine particles to those completely covering the solution. Using the proposed method, the solubility status can be automatically screened in real time by using a display and camera on a tablet or mobile phone. Therefore, by combining an automatic solubility changing system with the proposed method, a fully automated process could be achieved without human intervention.

Super-Linear-Threshold-Switching Selector with Multiple Jar-Shaped Cu-Filaments in the Amorphous Ge3 Se7 Resistive Switching Layer in a Cross-Point Synaptic Memristor Array.

The learning and inference efficiencies of an artificial neural network represented by a cross-point synaptic memristor array can be achieved using a selector, with high selectivity (Ion /Ioff ) and sufficient death region, stacked vertically on a synaptic memristor. This can prevent a sneak current in the memristor array. A selector with multiple jar-shaped conductive Cu filaments in the resistive switching layer is precisely fabricated by designing the Cu ion concentration depth profile of the CuGeSe layer as a filament source, TiN diffusion barrier layer, and Ge3 Se7 switching layer. The selector performs super-linear-threshold-switching with a selectivity of > 107 , death region of -0.70-0.65 V, holding time of 300 ns, switching speed of 25 ns, and endurance cycle of > 106 . In addition, the mechanism of switching is proven by the formation of conductive Cu filaments between the CuGeSe and Ge3 Se7 layers under a positive bias on the top Pt electrode and an automatic rupture of the filaments after the holding time. Particularly, a spiking deep neural network using the designed one-selector-one-memory cross-point array improves the Modified National Institute of Standards and Technology classification accuracy by ≈3.8% by eliminating the sneak current in the cross-point array during the inference process.

RNN-Based On-Line Continuous Gait Phase Estimation from Shank-Mounted IMUs to Control Ankle Exoskeletons.

Several research groups have developed and studied powered ankle exoskeletons to improve energetics of healthy subjects and the mobility of elderly subjects, or to reduce asymmetry in gaits induced by strokes. To achieve optimal effect, the timing of assistive torque has been proved to be of crucial importance. Previous studies estimated the onset timings mostly by extrapolating the time horizon from past gait events observed with sensors. Such methods have inherently limited performance when subjects are not walking at steady frequencies. To overcome such limitation and allow the use of exoskeletons in various scenarios in a daily life, we propose to estimate the gait phase as a continuous variable progressing over a gait cycle, hence allowing immediate response to frequency changes rather than iteratively correcting it after each cycle. Our method uses recurrent neural networks to estimate gait phases out of an inertial measurement unit (IMU) every 10 ms. By replacing foot sensors with an IMU we can obtain rich enough information to estimate gait phase continuously as well as avoid physical damage in sensors from ground impacts. Our preliminary tests with 2 healthy subjects showed qualitatively positive outcomes regarding the gait phase estimation and the assistive torque control.

Compact Hip-Force Sensor for a Gait-Assistance Exoskeleton System.

In this paper, we propose a compact force sensor system for a hip-mounted exoskeleton for seniors with difficulties in walking due to muscle weakness. It senses and monitors the delivered force and power of the exoskeleton for motion control and taking urgent safety action. Two FSR (force-sensitive resistors) sensors are used to measure the assistance force when the user is walking. The sensor system directly measures the interaction force between the exoskeleton and the lower limb of the user instead of a previously reported force-sensing method, which estimated the hip assistance force from the current of the motor and lookup tables. Furthermore, the sensor system has the advantage of generating torque in the walking-assistant actuator based on directly measuring the hip-assistance force. Thus, the gait-assistance exoskeleton system can control the delivered power and torque to the user. The force sensing structure is designed to decouple the force caused by hip motion from other directional forces to the sensor so as to only measure that force. We confirmed that the hip-assistance force could be measured with the proposed prototype compact force sensor attached to a thigh frame through an experiment with a real system.

Effects of assistance timing on metabolic cost, assistance power, and gait parameters for a hip-type exoskeleton.

There are many important factors in developing an exoskeleton for assisting human locomotion. For example, the weight should be sufficiently light, the assist torque should be high enough to assist joint motion, and the assistance timing should be just right. Understanding how these design parameters affect overall performance of a complex human-machine system is critical for the development of these types of systems. The present study introduces an assistance timing controller that regulates assistance timing such that peak joint velocity and peak assistance power are offset by a reference value for our hip-type exoskeleton. This is followed by measuring the manner in which various assistance timing references affect an important metric for performance, namely metabolic cost. The results indicate that net metabolic cost exhibits a concave up pattern with the most reduction of 21%, when compared to walking without the exoskeleton, at 0% assistance timing reference. The study also examines assistance timing's effect on gait parameters; increase in assistance timing reference increases step length, decreases cadence, and increases walk ratio (i.e. step length/cadence ratio) during treadmill walking.

A Wearable Virtual Chair with the Passive Stability Assist.

This paper introduces a wearable device which performs function of swinging chair with worn status on the legs. The users with the proposed device can sit in anyplace and experience the stable swing motion. The device is designed to maintain the stability within the stable swing region while moving back and forth by external forces or user intension. The coupled motion between ankle and knee provides the users concave swing motion in chair mode, while the joints passively follows the motion of the legs in normal gait mode. The key feature of this stable motion is a CAM-drive implemented around the ankle frame and connected to the knee joint by wires. With any directional motion of the ankle joint, the knee joint rotate only one direction to lift up the body of the user. So it can move following concave equilibrium line. We verified the payload of the device is more than 70 kg in computer-aided stress simulation as well as in experiments.

Surgical robot for single-incision laparoscopic surgery.

This paper introduces a novel surgical robot for single-incision laparoscopic surgeries. The robot system includes the cone-type remote center-of-motion (RCM) mechanism and two articulated instruments having a flexible linkage-driven elbow. The RCM mechanism, which has two revolute joints and one prismatic joint, is designed to maintain a stationary point at the apex of the cone shape. By placing the stationary point on the incision area, the mechanism allows a surgical instrument to explore the abdominal area through a small incision point. The instruments have six articulated joints, including an elbow pitch joint, which make the triangulation position for the surgery possible inside of the abdominal area. The presented elbow pitch structure is similar to the slider-crank mechanism but the connecting rod is composed of a flexible leaf spring for high payload and small looseness error. We verified the payload of the robot is more than 10 N and described preliminary experiments on peg transfer and suture motion by using the proposed surgical robot.