The progress observed in ‘soft robotics’ brought some promising research in flexible tactile, pressure and force sensors, which can be based on polymeric composite materials. Therefore, in this paper, we intend to evaluate the characteristics of a force-sensitive material—polyethylene-carbon composite (Velostat^®) by implementing this material into the design of the flexible tactile sensor. We have explored several possibilities to measure the electrical signal and assessed the mechanical and time-dependent properties of this tactile sensor. The response of the sensor was evaluated by performing tests in static, long-term load and cyclic modes. Experimental results of loading cycle measurements revealed the hysteresis and nonlinear properties of the sensor. The transverse resolution of the sensor was defined by measuring the response of the sensor at different distances from the loaded point. Obtained dependencies of the sensor’s sensitivity, hysteresis, response time, transversal resolution and deformation on applied compressive force promise a practical possibility to use the polyethylene-carbon composite as a sensitive material for sensors with a single electrode pair or its matrix. The results received from experimental research have defined the area of the possible implementation of the sensor based on a composite material—Velostat^®.

This review is dedicated to the development of molecularly imprinted polymers (MIPs) and the application of MIPs in sensor design. MIP-based biological recognition parts can replace receptors or antibodies, which are rather expensive. Conducting polymers show unique properties that are applicable in sensor design. Therefore, MIP-based conducting polymers, including polypyrrole, polythiophene, poly(3,4-ethylenedioxythiophene), polyaniline and ortho-phenylenediamine are frequently applied in sensor design. Some other materials that can be molecularly imprinted are also overviewed in this review. Among many imprintable materials conducting polymer, polypyrrole is one of the most suitable for molecular imprinting of various targets ranging from small organics up to rather large proteins. Some attention in this review is dedicated to overview methods applied to design MIP-based sensing structures. Some attention is dedicated to the physicochemical methods applied for the transduction of analytical signals. Expected new trends and horizons in the application of MIP-based structures are also discussed.

Human falls pose a serious threat to the person’s health, especially for the elderly and disease-impacted people. Early detection of involuntary human gait change can indicate a forthcoming fall. Therefore, human body fall warning can help avoid falls and their caused injuries for the skeleton and joints. A simple and easy-to-use fall detection system based on gait analysis can be very helpful, especially if sensors of this system are implemented inside the shoes without causing a sensible discomfort for the user. We created a methodology for the fall prediction using three specially designed Velostat^®-based wearable feet sensors installed in the shoe lining. Measured pressure distribution of the feet allows the analysis of the gait by evaluating the main parameters: stepping rhythm, size of the step, weight distribution between heel and foot, and timing of the gait phases. The proposed method was evaluated by recording normal gait and simulated abnormal gait of subjects. The obtained results show the efficiency of the proposed method: the accuracy of abnormal gait detection reached up to 94%. In this way, it becomes possible to predict the fall in the early stage or avoid gait discoordination and warn the subject or helping companion person.

A Digital twin (DT) enables the acquisition and subsequent analysis of large amounts of process data. Various machine learning (ML) algorithms exist for analysis and prediction that can be used in this scenario. However, there is very little understanding of the relative merit of these methods. This paper proposes a DT framework in the context of predictive maintenance in smart manufacturing to compare the prediction efficacy of prevalent ML models. Data-driven models were developed using machine learning algorithms to predict surface roughness and power consumption during a CNC turning operation. Three process parameters, namely cutting velocity, feed rate, and depth of cut, and two dependent parameters, surface roughness and power consumption, were selected for model development. Seven ML algorithms were tested for each response parameter: Linear Regression, XGB Regressor, Random Forest Regressor, Average Ensemble, AdaBoost Regressor, SVR, and MLP. The results of the comparative analysis of the ML algorithms showed that the Random Forest Regressor is the best prediction model for surface roughness, with the highest R² (94.2% ± 2.4%), lowest MAE (0.011 ± 0.002), lowest MAPE (15.6% ± 4.0%), and lowest RMSE (0.017 ± 0.003), while the XGB Regressor demonstrated the best performance for power consumption prediction, with the highest R² (98.9% ± 0.5%), lowest MAE (22.513 ± 4.424), lowest MAPE (3.0% ± 0.5%), and lowest RMSE (42.650 ± 8.933). The best-performing machine learning algorithm was subsequently utilized in the data-driven models, helping to achieve an improved surface finish. This enables predictive maintenance, reducing energy usage for more sustainable production.