Recent advances in technology have led to the development of various intelligent sensor systems, in which pattern recognition and sensor fusion algorithms play a crucial role in most of the cases. For effective operation of these algorithms, advanced solutions are required in many areas, such as pre-processing, feature extraction, feature selection, classification, decision making, state estimation, implementation, etc.

Intelligent sensor systems can be based on signals of various sensor types. Many applications use sensors which provide 2D or 3D data, such as cameras and LiDARs, where computer vision solutions are required. Others apply time-series analysis on signals collected from acoustic sensors, inertial sensors (IMU), magnetometers, etc. Both types or their fusion are widely used in industrial, medical, health, and entertainment applications (e.g., robotics, pose estimation, human–computer interaction, navigation, intelligent transportation systems, activity, and movement analysis, etc.).

The processing can be performed on a central unit or decentralized, where the required computations are done on the embedded system of the units. Since most of the applications require real-time operation, the design of these pattern recognition and sensor fusion algorithms and their implementation on the embedded systems are challenging tasks.

Intelligent sensor fusion solutions and autonomous robots are in synergy, since the latter task is realized successfully only if the former provides appropriate outputs. Intelligent sensors aim to establish robust solutions for measuring physical quantities, such as linear acceleration, 3D orientation and/or actuator torque (current). Robustness is addressed with importance, since sensors provide data with errors coming from manufacturing, noise and external disturbances. Therefore, intelligent solutions (e.g., machine learning tools and sophisticated filtration techniques) are formulated with the objective to fuse multiple sensors and provide data with enhanced reliability and accuracy. An example of this is the localization of robot systems, where IMU, encoder, magnetometer, LiDAR and GPS measurements are incorporated in intelligent sensor solutions with the aim to obtain the instantaneous robot pose. Autonomous robots utilize the information provided by intelligent sensors to obtain paths between start and destination coordinates. These systems utilize sophisticated path planner algorithms, which provide optimal trajectories with obstacle avoidance. Moreover, intelligent controller algorithms contribute to successful autonomous behavior. These algorithms are designed with the help of artificial intelligence tools to obtain both trajectory tracking and optimized closed loop performance. An example of this is fuzzy logic augmented nonlinear low-level controllers, which accomplish unique control requirements, such as limited current transients and jerks during the motion of the system.

The ARIS research group’s work is focused on the aforementioned tasks. In the field of intelligent sensors, we aim to develop

• Specialized sensor hardware and software solutions for dedicated tasks
• Specialized measurement setups
• Novel methods for offline and online calibration of different sensors
• Signal analysis and pattern recognition-based intelligent and AI augmented methods
• Low cost and low computational requirements-based algorithms for embedded systems

The autonomous robots field addresses the topics listed as follows.

• Novel sensor fusion solutions for reliable localization of mobile robots
• Development of adaptive algorithms for the insurance of robust (changing environment independent) pose estimation
• Analysis of environmental characteristics of mobile robots and it's incorporation in sensor fusion solutions
• Low cost, AI-based algorithms for simultaneous localization and mapping
• Cooperation-based methods for multi-robot systems
• Advanced control solutions for mobile robots

Our aims are further strengthened by interest in both university collaborations (e.g., between research groups) and industrial applications/projects.

Leaders:

Dr. Akos Odry

Dr. Peter Sarcevic

Members:

Prof. Dr. Jozsef Sarosi

Dominik Csik

Massimo Stefanoni

Richard Pesti

Dr. Vladimir Tadic (external)

Dr. Istvan Kecskes (external)

Dr. Juvenal Rodriguez (external)

Most important publications:

Á. Odry, R. Fullér, I. J. Rudas and P. Odry, "Kalman filter for mobile-robot attitude estimation: Novel optimized and adaptive solutions," Mechanical Systems and Signal Processing, vol. 110, pp. 569-589, 2018

P. Sarcevic, Z. Kincses and S. Pletl, "Online human movement classification using wrist-worn wireless sensors," Journal of Ambient Intelligence and Humanized Computing, vol. 10, no. 1, pp. 89-106, 2019.

Á. Odry, I. Kecskes, P. Sarcevic, Z. Vizvari, A. Toth, and P. Odry, "A Novel Fuzzy-adaptive Extended Kalman Filter for Real-time Attitude Estimation of Mobile Robots," Sensors, vol. 20, no. 3, 803, 2020

Á. Odry, "An Open-Source Test Environment for Effective Development of MARG-Based Algorithms," Sensors, vol. 21, no. 4, 1183, 2021

I. Kecskés, A. Odry, V. Tadić and P. Odry, "Simultaneous calibration of a hexapod robot and an IMU sensor model based on raw measurements," IEEE Sensors Journal, vol. 21, no. 13, pp. 14887-14898, 2021.

Special Issues:

"Modeling, Sensor Fusion and Control Techniques in Applied Robotic" in Machines

"Pattern Recognition and Sensor Fusion Solutions in Intelligent Sensor Systems" in Electronics

Projects:

Title: Reliable localization of mobile robots based on intelligent sensor fusion methods

Leader: Dr. Akos Odry

Number and source: 142790, FK_22 funding scheme provided by the National Research, Development and Innovation Fund of Hungary

Duration: 48 (2022-09-01 - 2026-08-31)

Title: Presentation of novel sensor fusion and control algorithms in robotics

Leader: Dr. Akos Odry

Number and source: 141384, MEC_R funding scheme provided by the National Research, Development and Innovation Fund of Hungary

Duration: 11 (2022-01-01 - 2022-12-31)

Budget: 1.301 Million HUF

Collaborations/Partners:

• Laboratory for Information Technologies, Faculty of Electrical Engineering, University of Ljubljana, Ljubljana, Slovenia. Lead: Prof. Dr. Sašo Tomažič
• Robotics, Navigation, and Control Systems Laboratory (RNCSL), Department of Mechanical and Aerospace Engineering, Carleton University, Ottawa, Ontario, Canada. Lead: Hashim A. Hashim, Ph.D., P.Eng., SMIEEE
• Energy and Sustainability Laboratory, Faculty of Engineering, Autonomous University of Queretaro, Mexico. Lead: Prof. Dr. Juvenal Rodríguez
  
• Department of Mechanical, Energy and Management Engineering, University of Calabria, Rende, Italy‎. Lead: Prof. Dr. Giuseppe Carbone