It is pivotal to monitor and examine the plant disease during in situ measurements to abate the crop loss. For this purpose, leaf wetness sensors (LWS) are widely used. However, for the LWS during in situ measurements, operational exposure is always a concern considering the plant growth at different stages. During the plant growth, the stem angle changes and even the leaf canopy bends either inward or outward due to environmental factors or physical trauma. Thus, LWS placed on the leaf canopy may produce erroneous results. In this letter, we have examined the effect of leaf bending radius (outward or inward) and angle (from 0° to 90°) on the flexible LWS fabricated on the polyamide substrates. LWS comprises of interdigitated electrodes (IDEs) having interelectrode spacing 0.05 cm. Fabricated LWS are 3.5 cm long and 1.5 cm wide in dimension. We have used the two LWS viz. one bare IDEs and another with molybdenum disulfide (MoS ₂ ) coated LWS. Lab experiments indicated that sufficient wetness remained on the bare IDEs and MoS ₂ -coated IDEs till 40° and 70° of bending angle, respectively. Subsequently, when the LWS are bended outward or inward, bare IDEs and MoS ₂ -coated IDEs retain water molecules till 0.7 and 1 cm, respectively, when bended from its initial length (3.5 cm).

Electrochemical motion sensors are widely used in the creation of seismology devices. The key element of an electrochemical motion sensor is a miniature electrochemical cell with platinum electrodes, the current, in which changes under the action of external mechanical signals. This work explores the possibility of replacing platinum with much cheaper carbon. Two types of configurations of the converting element have been studied, in one of which the electrodes are located on the walls of a narrow channel, in which liquid flows under the influence of an external signal. In the second configuration, electrodes are made on the sides of the plate, and microscopic holes are made in the plate to transfer fluid between the electrodes. Comparison of the sensitivity of sensors with platinum and carbon electrodes showed the similarity of their characteristics, provided that the electrode systems are similar in their geometry.

Near-field radio sensors can be implemented both as wearables for continuous monitoring and as furnishing attachments for covert operation, without requiring skin touch, enabling convenient deployment and high user comfort. The sensor can output cardiac and respiratory waveforms, systemic pulses, muscle contraction, eye movement, and tissue vibration for various biomedical, biological, and cyberphysical applications. The radio signal can go through layers of fabrics or animal surface coverings with little loss or dispersion to couple directly inside the living body. We will illustrate the theory, design variations, and verified results, and discuss present limitations and outlooks for technology adoption in practical scenarios.

The acquisition of machine learning (ML) datasets by meas-urements for automotive radar data requires many resources and time. On the other hand, the simulation of complex traffic environments with sufficient level-of-detail is challenging, too. In this letter, a middle way is proposed in which real measurements are virtually augmented by reflections obtained from simulation models. The augmented measurements are then replayed with a hardware-based radar target simulator (RTS). This enables the fast creation of application-specific datasets, as well as advanced functionality tests of algorithms at deployment. To demonstrate the efficacy of the virtually augmented radar data, a set of test drives is augmented by virtual pedestrians performing traffic gestures. Then, a classifier trained on the created dataset is demonstrated to achieve a high classification accuracy of 84.0% on real test data.

In recent years, sensing systems have been extensively used for motion detection, activity detection, and gesture recognition, among a few other important applications. Wearable sensors, such as smartwatches and smartphones, contain accelerometers and gyroscope sensors that sense a user's movements and activities to detect abnormal events. Inspired by recent breakthroughs in neural machine translation and the generation of image descriptions, we propose a first-of-its-kind novel attention-based encoder–decoder model to generate a caption to summarize various activities detected for a period from smartphone sensor data. The proposed model architecture consists of three layers: 1) the bidirectional long short-term memory (BiLSTM) incorporates both past and future information from the raw sensor data, then generates features; 2) an attention mechanism is used to assign different weights depending on the feature importance; and 3) an LSTM layer is used generate a sequence of activities, and then, a caption generator module is used to generate the caption. The performance of the proposed model is evaluated using two widely used public datasets (UCI-HAR and WISDM) and one experimental dataset. The model achieves good accuracy on UCI-HAR and our experimental dataset compared to WISDM dataset. The proposed model is able to achieve an average word error rate of 8.20%, accuracy of 90.75% with the UCI-HAR dataset, and an average word error rate of 10%, as well as accuracy of 90% with our experimental dataset.

Field robots require autonomy in diverse environments to navigate and map their surround-ings efficiently. However, the lack of diverse and comprehensive datasets hinders the evaluation and development of autonomous field robots. To address this challenge, we present a multimodal, multisession, and diverse terrain dataset for the ground mapping of field robots. First of all, we utilize a quadrupedal robot as a base platform to collect the dataset. Also, the dataset includes various terrain types, such as sandy roads, vegetation, and sloping terrain. It comprises RGB-D camera for ground, RGB camera, thermal camera, light detection and ranging (LiDAR), inertial measurement unit (IMU), and global positioning system (GPS). In addition, we provide not only the reference trajectories of each dataset but also the global map by leveraging LiDAR-based simultaneous localization and mapping (SLAM) algorithms. Also, we assess our dataset from a terrain perspective and generate the fusion maps, such as thermal-LiDAR and RGB-LiDAR maps to exploit the information beyond the visible spectrum.

Connected autonomous vehicles (CAVs) are currently in the testing and evaluation stage as there are still many challenges to be addressed before CAVs can appear on public roads at mass scale. This letter outlines key areas needed for improvement in CAV sensors technology to accelerate deployment of CAVs. These areas include low-fidelity sensor models used in virtual testing, definition of performance envelopes for individual or fused CAV sensors, impact of sensor deterioration on a CAV's decision-making system, stimulation of real sensors with data generated in virtual testing environments, and a shortfall in industry standards for CAV perception systems testing.

With the rapid development of science and technology, traditional human–computer interaction methods are evolving in various directions. Specifically, somatic interaction technology plays an important role in education, entertainment, augmented reality, virtual reality, and industry. To enhance the immersion and realism of interacting with remote or virtual environments, this letter proposes a data glove capable of multimodal feedback with vibration and thermal tactile feedback and tracking the user's hand posture. In addition, the Unity engine is utilized to build some simple scenes for better testing and display. The system effectively improves the user's immersion and realism in virtual space, despite some limitations of its large size and weight.