Experimental Evaluation of Navigation Slam and Line Follower on Automated Guided Vehicles to Improve Industrial Operational Flexibility

Figure Ilustration AI

FORMOSA NEWS - Jakarta - SLAM vs Line Follower: How Smart Autonomous Guided Vehicles Slash Warehouse Reconfiguration Time by 87 Percent. Industrial automation researchers at Universitas Global Jakarta have published a comprehensive experimental evaluation comparing Simultaneous Localization and Mapping (SLAM) navigation with traditional line-follower systems on Automated Guided Vehicles (AGVs). Led by Ariep Jaenul, Rio Handoko, and Brainvendra Widi Dionova, the 2026 study reveals that while classic line-following systems remain faster and more precise on fixed, predictable paths, LiDAR-based SLAM systems slash route reconfiguration time by 87.8% and maintain complete task completion when facing physical obstacles. Published in the Formosa Journal of Computer and Information Science, these findings offer manufacturing leaders and logistics planners a data-driven blueprint for choosing the right robotic navigation technology to balance operational speed against shop-floor adaptability.

The Challenge of Flexibility in Smart Manufacturing

As Industry 4.0 shifts factory floors toward agile manufacturing, traditional automated material handling faces a critical bottleneck. Traditional AGVs rely heavily on line-follower technology, navigating along magnetic tape, painted lines, or optical tracks embedded in the floor. While these systems are cost-effective and highly reliable, every physical modification to a factory layout demands manual re-taping and track re-installation—costly interventions that interrupt production schedules. Autonomous mobile robots equipped with SLAM solve this issue by using laser scanners and inertial sensors to map their surroundings in real time, enabling markerless navigation. However, smart navigation adds computational complexity and potential position drift. To determine whether SLAM's operational flexibility outweighs the raw precision of physical line-following, the Jakarta research team designed a multi-scenario experimental evaluation using a single physical AGV platform.

Experimental Setup and Methodology
The researchers conducted physical tests at the PT Akebono Brake Astra Indonesia facility, utilizing a 10-meter base trajectory. A single AGV unit was modified to host both navigation modes, eliminating structural vehicle bias.

• Hardware & Architecture: The SLAM architecture integrated an RPLiDAR A2M12 $360^\circ$ laser scanner and an MPU-6050 Inertial Measurement Unit (IMU). Sensor data was processed by an Intel NUC computer running Ubuntu and the Robot Operating System (ROS) with Hector SLAM, while an Arduino Mega 2560 executed direct BLDC motor control.
  
• Test Scenarios: Both systems were systematically evaluated across four operational variables:
  
• Speed Variations: Tested at $0.188\text{ m/s}$ and $0.377\text{ m/s}$ over a straight 10-meter path.
  
• Obstacle Challenges: Evaluated against 1 to 10 physical barriers placed along the trajectory.
  
• Payload Changes: Tested across 10 mass levels ranging from $50.08\text{ kg}$ to $200.32\text{ kg}$.
  
• Route Reconfiguration: Evaluated across five turn offset variations ($40\text{ cm}$ to $200\text{ cm}$) to measure setup time and travel performance.
  

Key Findings: Precision vs. Adaptability
The experimental results highlight a clear trade-off between straight-line efficiency and dynamic flexibility:

• Superior Line Precision on Static Routes: On unobstructed paths, line-follower navigation delivered higher accuracy and faster travel times. At $0.188\text{ m/s}$, the line follower achieved a minimal endpoint position error of $1.06\text{ cm}$ compared to $8.42\text{ cm}$ for SLAM. At $0.377\text{ m/s}$, line-follower error rose to $2.25\text{ cm}$, whereas SLAM error remained stable at $8.49\text{ cm}$.
  
• Uninterrupted Obstacle Avoidance: When encountering physical barriers, line-follower AGVs registered $100\%$ task failure, coming to a full stop without alternative options. SLAM navigation successfully rerouted around barriers in $100\%$ of test runs, completing the 10-obstacle course with only a minor travel time increase from $45.32\text{ seconds}$ to $51.12\text{ seconds}$.Payload Stability: Under varying payloads ($50.08\text{ kg}$ to $200.32\text{ kg}$), the line follower saw its position error drift by $71.6\%$ (from $4.05\text{ cm}$ to $6.95\text{ cm}$). In contrast, SLAM demonstrated higher consistency, fluctuating by only $7.1\%$ (from $9.13\text{ cm}$ to $9.78\text{ cm}$) due to continuous sensor feedback.
  
• 87.8% Faster Route Reconfiguration: Changing a turning layout required an average of $286\text{ seconds}$ for line-follower systems due to physical tape removal and re-application. SLAM required just $35\text{ seconds}$ through Human-Machine Interface (HMI) software updates—making route changes $8.2\text{ times}$ faster.
  

Real-World Impact and Industry Adoption
For modern manufacturing facilities, these results demonstrate that choosing an AGV navigation system depends on layout permanence. While line followers remain ideal for high-volume, repetitive transfer tasks on permanently fixed paths, SLAM technology provides essential ROI for agile, multi-product production environments where layouts change frequently. Factory personnel confirmed this operational value in survey responses, awarding SLAM navigation an overall composite rating of $4.44$ out of $5$. Operational benefits and safety performance scored highest ($4.69$ and $4.63$), though lower scores for ease of operation ($3.94$) indicate that software interface design remains a key area for user training.

Author Profile
Ariep Jaenul, M.T. – Universitas Global Jakarta, Department of Industrial Engineering. Specialist in industrial robotics, autonomous navigation systems, and automated material handling.
Rio Handoko, M.T. – Universitas Global Jakarta, Faculty of Engineering and Computer Science. Researcher in mechatronics, embedded systems, and industrial control systems.
Brainvendra Widi Dionova, M.T. – Universitas Global Jakarta, Department of Computer Engineering. Expert in computer vision, robotics software architecture, and sensor fusion.

Source
Ariep Jaenul, Rio Handoko, Brainvendra Widi Dionova. Experimental Evaluation of Navigation Slam and Line Follower on Automated Guided Vehicles to Improve Industrial Operational Flexibility. Formosa Journal of Computer and Information Science (FJCIS). Vol. 5, No. 2, Halaman 401–416)
DOI: https://doi.org/10.55927/fjcis.v5i2.17278
URL: https://journal.formosapublisher.org/index.php/fjcis