Driving style can be characteristically divided into normal and aggressive. Related researches show that useful information about driving style can be extracted using vehicle's inertial measurement signals with the help of GPS. However, for public transportation the GPS sensor isn't necessary because of repetition of the route. This assumption helps to create low-cost intelligent public transport monitoring system that is capable to classify aggressive and normal driver. In this paper, we propose pattern recognition approach to classify driving style into aggressive or normal automatically without expert evaluation and knowledge using accelerometer data when driving the same route in different driving styles. 3-axis accelerometer signal statistical features were used as classifier inputs. The results show that aggressive and normal driving style classification of 100% precision is achieved using collected data when driving the same route.Index Terms-Vehicle driving, intelligent vehicles, pattern recognition, accelerometer.

This paper presents a passengers counting system based on computer vision. System prototype were created and installed in Kaunas public city transport. Four algorithms were created to calculate passengers on public transport and their advantages and disadvantages were analyzed. Qualitative detection algorithms analysis carried out. Promising results were obtained with the Algorithm of barrier simulation for zones (ABSZ) which has low false rate and it is effective for people-counting. Counting results information can be used for public transport optimization or service quality improvement.

The main goal of this work is the development of Arduino microcontroller based integral solution for the smart, assistive mobility hardware. The paper presents schematics of device control and electrics, the control logic, the power regulation chain model and affiliated calculate efficiency parameters. The overall experimental investigation of real-life performance of our Arduino model proved quite successful. The added power regulator allowed achieving the stable outputs (±0.001V) to mimic the original controller. The deviation in trajectory compared to the ideal model was small (< 10 %) and can be further reduced by doing micro corrections.

The paper describes our implementation and experimental evaluation of a touch surface control algorithms developed for the smart mobile furniture. We begin the article by presenting the working principle of touch surface, followed by the descriptions of proposed eyes-free control algorithms aimed to avoid random touches. The experimental evolution and analysis shown the problems affiliated with the usage of standard hardware by manipulating it by nose and foot finger (compared vs a hand finger), and allowed the initial determination of recognition accuracy, performance (in time) and overall user rating.

No abstract