TEM Journal. Volume 11, Issue 1, pages 30-36, ISSN 2217-8309, DOI: 10.18421/TEM111-04, February 2022. 30 TEM Journal – Volume 11 / Number 1 / 2022. Linear Differential Driven Wheel Mobile Robot Based on MPU9250 and Optical Encoder Ilham Rustam 1 , Nooritawati Md Tahir 2,3 , Ahmad Ihsan Mohd Yassin 2 , Nurbaiti Wahid 1 , Abdul Hafiz Kassim 1 1 School of Electrical Engineering Department, College of Engineering, Universiti Teknologi MARA, Terengganu, Malaysia 2 School of Electrical Engineering Department, College of Engineering, Universiti Teknologi MARA, Selangor, Malaysia 3 Institute for Big Data Analytics and Artificial Intelligence (IBDAAI), Universiti Teknologi MARA, Selangor, Malaysia Abstract – In this paper, the operability of a custom- built differential drive wheel mobile robot was evaluated using a self-driving algorithm in practical application. Several key parameters need to be realized by the WMR to enable succinct self-driving algorithm evaluations and parameter derivations. The main components of the WMR development were examined in this work, specifically parameter derivations, hardware implementation, steering algorithm and WMR operability. The result obtained indicate that with proper design and careful calibration of sensors, the required driving parameters could be realized. This enables the WMR to potentially serve as data collection platform for future evaluation of self-driving algorithms. Keywords – self-driving algorithm, autonomous vehicles, sensor fusion, inertial measurement unit, wheel encoders. DOI: 10.18421/TEM111-04 https://doi.org/10.18421/TEM111-04 Corresponding author: Ilham Rustam, School of Electrical Engineering Department, College of Engineering, Universiti Teknologi MARA, Malaysia. Email: ilhamr480@uitm.edu.my Received: 02 October 2021. Revised: 11 January 2022. Accepted: 17 January 2022. Published: 28 February 2022. © 2022 Ilham Rustam et al; published by UIKTEN. This work is licensed under the Creative Commons Attribution‐NonCommercial‐NoDerivs 4.0 License. The article is published with Open Access at www.temjournal.com 1. Introduction Amongst the most popular platform used to study the practical applications of artificial intelligence (AI) algorithms are the robotics platforms. The robotics platforms can be divided into two classifications that is fixed or mobile [1]. As opposed to fixed robot, which mainly can be found in industrial settings, mobile robot offers much more dynamic and exhaustive computational problems, perfectly suited to challenge the AI algorithm computational capabilities. Consequently, wheeled mobile robot (WMR) was found to be one of the known being used due to its cost-effective nature that offers real world complexity in its problem. Within the research context, tracking a path trajectory, obstacle avoidance and point stabilization are amongst the most elaborated control problems to be solved in recent years. Moreover, path trajectory is the most popular due to the high nonlinearity nature and coupling property of the problem [2]. In addition, several notable works have successfully solved path tracking control problem of WMR using techniques such as sliding mode control [3], adaptive control [4], and backstepping control [5]. However, it should be noted that most of the work assumed that ideally there would be no wheel slippage or skidding occurrences during vehicle movements [6]. While this assumption is feasible during simulation, unfortunately this is not true in real world scenario or in practical applications. Within the context of optimal path tracking and object avoidance of the WMR, wheel slips and skids could cause significant reading errors to WMR positioning and orientation [7], hence these issues need to be addressed appropriately. Some of the factors that caused wheel slips and skids include tire deformation, weak friction between road and the