IAES International Journal of Robotics and Automation (IJRA)

Vol. 13, No. 2, June 2024, pp. 131~139

ISSN: 2722-2586, DOI: 10.11591/ijra.v13i2.pp131-139  131

Robot navigation on inclined terrain using social force model

Muhammad Fariz Daffa1, Bima Sena Bayu Dewantara2, Setiawardhana2

1
Department of Electrical Engineering, Politeknik Elektronika Negeri Surabaya, Surabaya, Indonesia
2
Department of Informatics and Computer Engineering, Politeknik Elektronika Negeri Surabaya, Surabaya, Indonesia

Article Info ABSTRACT

Article history: This research introduces an innovative approach to address the limitations of
the commonly used social force model-based robot navigation method on
Received Jun 1, 2023 flat terrain when applied to sloped terrain. The incline of the terrain becomes
Revised Feb 20, 2024 a crucial factor in calculating the robot’s steering output when navigating
Accepted Feb 26, 2024 from the initial position to the target position while avoiding obstacles.
Therefore, we propose a social forced model-based robot navigation system
that can adapt to inclined terrain using inertial measurement unit sensor
Keywords: assistance. The system can detect the surface incline in real time and
dynamically adjust friction and gravitational forces, ensuring the robot’s
Inclined terrain speed and heading direction are maintained. Simulation results conducted
Inertial measurement unit using CoppeliaSim show a significant improvement in speed adjustment
Robot navigation efficiency. With this new navigation system, the robot can reach its
Simulation destination in 59.935089 seconds, compared to the conventional social
Social force model forced model which takes 63.506442 seconds, the robot is also able to
reduce slip to reduce wasted movement. This method shows the potential of
implementing a faster and more efficient navigation system in the context of
inclined terrain.
This is an open access article under the CC BY-SA license.

Corresponding Author:
Bima Sena Bayu Dewantara
Department of Informatics and Computer Engineering, Politeknik Elektronika Negeri Surabaya
Surabaya, East Java, Indonesia
Email: [email protected]

1. INTRODUCTION
Terrain-based adaptive control has the primary goal of assisting robots in undergoing stable and safe
movement, especially when operating on different types of terrain that may differ in characteristics.
Achieving this goal requires the robot’s ability to accurately identify surface slopes and adapt its actions and
responses according to the terrain conditions encountered [1]–[3]. One of the key aspects of this adaptability
is the ability to regulate the forces acting on the robot so that changes in speed or direction of movement do
not occur suddenly.
In this research framework, the main focus is on developing an adaptive social force model (SFM)-
based navigation model, especially in the context of inclined terrain. Previously, social force navigation
models have been successfully used in various applications, such as pedestrian avoidance [4]–[10],
healthcare robots [11] drones [12], [13], evacuation robots [14]–[17], and navigation of soccer robots [18],
[19] some also modify the SFM [20]–[22]. However, in most cases, the use of these models is limited to flat
surfaces and does not consider changes that may occur to the robot during travel or navigation.
Therefore, this research aims to develop a SFM system that can adapt to various terrain conditions
that may be faced by the robot. This system will be equipped with the ability to detect changes in surface
slope in real time and the ability to adjust the forces acting on the robot according to the context of the terrain
being traveled. As a result, the acceleration, speed, and heading of the robot can be dynamically adjusted

Journal homepage: http://ijra.iaescore.com

132  ISSN: 2722-2586

according to the slope of the terrain, allowing the robot to maintain movement stability and respond more
effectively to terrain conditions. The forces generated by social force mode will be integrated with other
forces that affect the robot’s movement, including friction forces that may occur on the robot. Through the
integration of these forces, the system will ensure that the speed and heading obtained by the robot remain
stable and consistent in the face of diverse terrain conditions. Thus, the robot will have the ability to operate
efficiently and safely in a variety of inclined terrains [23]–[25].

2. METHOD
2.1. Social force model
SFM, introduced by Helbing and Molnar [26], [27], is a system used to predict the possible
behavioral-based movements of agents or individuals based on the attractive and repulsive forces acting on
them. It considers both physical and social factors that influence agent movements. In the SFM, three types
of forces influence agent movements, namely an attractive force toward the goal, a repulsive force against a
static obstacle (i.e., walls, buildings, roads), and a repulsive force against a dynamic obstacle (i.e., human).
Figure 1 depicts the relation between all force components in the SFM framework.

Figure 1. The relation between all force components in SFM framework

The primary objective of the SFM is to determine the navigation force, denoted as 𝐹𝑛𝑎𝑣 , essential for
guiding a mobile robot through its environment. This is achieved through the intricate calculation of a
resultant force derived from three fundamental components. The first component involves an attractive force
directed towards the predefined goal, denoted as 𝐹𝑔 . Simultaneously, the second component introduces a
repulsive force aimed at mitigating potential collisions with static obstacles, represented by 𝐹𝑠 . Lastly, the
third component incorporates a repulsive force designed to counteract the influence of dynamic obstacles,
expressed as 𝐹𝑑 . The formulation of these components is fundamental to achieving a nuanced and balanced
navigation force, enabling the mobile robot to navigate effectively by harmonizing attraction towards the
goal and repulsion from obstacles, both static and dynamic, within its operational environment.

𝐹𝑛𝑎𝑣 = 𝐹𝑔 + 𝐹𝑠 + 𝐹𝑑 . (1)

𝐹𝑔 = 𝑚 . 𝑎 (2)

The repulsive force against a static obstacle, denoted as 𝐹𝑠 , is composed of two main components:
𝑠 𝑠
the social repulsion force denoted as 𝑓𝑠𝑜𝑐 and the physical repulsion force denoted as 𝑓𝑝ℎ𝑦 . The social repulsion
force arises from the interactions between the robot and its surroundings, including people or stationary
objects. This force captures the social aspect of the robot’s environment, reflecting its ability to navigate and
interact safely with individuals and objects. On the other hand, the physical repulsion force stems from the
tangible interactions between the robot and static objects, such as walls or stationary obstacles. This force is
crucial for preventing collisions and ensuring the robot’s physical integrity. By summing up these two forces,
the overall repulsion force against static obstacles is determined. This comprehensive approach allows the

IAES Int J Rob & Autom, Vol. 13, No. 2, June 2024: 131-139
IAES Int J Rob & Autom ISSN: 2722-2586  133

robot to navigate its environment effectively, taking into account both social considerations and physical
barriers, ultimately contributing to a more robust and adaptive robotic system.
𝑠
𝑟𝑅 − 𝑑𝑅
𝑠
𝑓𝑠𝑜𝑐 = 𝑘 𝑠 𝑒𝑥𝑝 ( ) + 𝑒𝑅𝑠 , (3)
𝛹𝑠

𝑠
𝑓𝑝ℎ𝑦 = 𝑘 𝑠 (𝑟𝑅 − 𝑑𝑅𝑠 ) 𝑒𝑅𝑠 , (4)

𝑠 𝑠
𝐹𝑠 = 𝑓𝑠𝑜𝑐 + 𝑓𝑝ℎ𝑦 , (5)

𝑠
In the context of the provided formulation, 𝑓𝑠𝑜𝑐 stands for the robot’s social repulsion force against
𝑠
static obstacles, while 𝑓𝑝ℎ𝑦 represents the robot’s physical repulsion force exerted on static obstacles. The
parameter 𝑟𝑅 corresponds to the radius of the robot’s interaction area, and 𝑑𝑅𝑠 denotes the distance from the
robot to the nearest static obstacle. The coefficient 𝑘 𝑠 serves as a gain factor that determines the magnitude of
feedback received subsequently. Furthermore, 𝛹 𝑠 denotes the effective distance value influencing the
repulsion force to navigate around dynamic obstacles, and 𝑒𝑅𝑠 is a vector indicating the direction from which
static obstacles approach. The repulsive force against a dynamic obstacle 𝐹𝑑 arises when there are moving
obstacles around the robot. It is also formed through the sum of the social repulsion force and the physical
repulsion force. To anticipate the movement of dynamic obstacles, in the SFM framework, a proxemic area
radius is given that indicates the possible movements that can occur next.

𝑑 𝑑
𝐹𝑑 = 𝑓𝑠𝑜𝑐 + 𝑓𝑝ℎ𝑦 , (6)

𝑑
𝑟𝑟𝑑 − 𝑑𝑅
𝑑
𝑓𝑠𝑜𝑐 = 𝑘 𝑑 𝑒𝑥𝑝 ( ) 𝑒𝑅𝑑 , (7)
𝛹𝑑

𝑑
𝑓𝑝ℎ𝑦 = 𝑘 𝑑 (𝑟𝑟𝑑 − 𝑑𝑅𝑑 ) 𝑒𝑅𝑑 , (8)

𝑟𝑟𝑑 = 𝑟𝑅 + 𝑑𝑅𝑑 , (9)

𝑑
In the outlined framework, 𝑓𝑠𝑜𝑐 represents the robot’s social repulsion force against dynamic
𝑑
obstacles, while 𝑓𝑝ℎ𝑦 characterizes the robot’s physical repulsion force applied to dynamic obstacles. The
parameter 𝑟𝑅 signifies the value of the radius defining the robot’s detection area, 𝑑𝑅𝑑 indicates the distance
from the robot to the nearest dynamic obstacle. The coefficient 𝑘 𝑑 functions as a gain factor determining the
degree of subsequent feedback. Additionally, 𝑒𝑅𝑑 represents a vector indicating the direction from which
dynamic obstacles approach, and 𝛹 𝑑 signifies the value determining the effective distance of the repulsive
force to navigate around dynamic obstacles. The parameter 𝑟𝑟𝑑 is derived from the sum of the radius of the
robot’s proxemics area and the radius of the dynamic obstacle area.

2.2. Inclined force

In this paper, we proposed to consider the inclined terrain factor as shown in Figure 2. Since the
robot sometimes needs to navigate in a traverse position on an inclined terrain, a slippery condition may
occur and make the robot shift to the side due to gravitation force. The first step in this calculation is to find
the value of the normal force (N) acting on the robot, which can be expressed in the following equation
related to the forces affecting the robot when traveling on an inclined road, to make the SFM better adapted
in its navigation.

Figure 2. Robot on inclined terrain

Robot navigation on inclined terrain using social force model (Muhammad Fariz Daffa)
134  ISSN: 2722-2586

Σ𝐹𝑦 = 0 (10)

𝑁 − 𝑊. 𝑐𝑜𝑠 𝜃 = 0 (11)

𝑁 = 𝑊. 𝑐𝑜𝑠 𝜃 (12)

In the context of the given formulation, Σ𝐹𝑥 represents the sum of forces acting on the robot along
the x-axis. This includes the weight force, determined by the product of the robot’s mass and gravity, and is
pivotal in understanding the dynamics of the robot’s motion. The angle 𝜃 introduced in the formulation
signifies the slope of the terrain, measured by the robot’s sensor during implementation. This angle enhances
the analysis by accounting for the inclination of the surface on which the robot operates. Calculating the
resultant force on the x-axis involves determining both the friction force and the parallel force. The friction
force arises from the interaction between the robot’s wheels or contact points and the surface, opposing
motion. The parallel force is a component of the weight force parallel to the terrain’s incline.

Σ𝐹𝑥 = 𝐹𝑝 − 𝐹𝑓 (13)

The parallel force 𝐹𝑝 is obtained by multiplying the robot’s weight force by 𝑠𝑖𝑛 𝜃, indicating that
the greater the inclination of the plane, the greater the parallel force generated. Meanwhile, the friction force
𝐹𝑓 is calculated by multiplying the friction coefficient 𝜇 by the normal force 𝑁, which has been calculated
previously. This coefficient of friction depends on the material of the plane being traversed and the material
of the robot wheel. In this study, the coefficient of friction is considered constant. In this paper, we ignored
the dynamic obstacle since it can be simplified and equated with an approach to static obstacles. The resultant
of the robot’s force on the x-axis when going through the inclined plane can be added to the SFM so the
robot can minimize the force that occurs on the inclined plane.

𝐹𝑝 = 𝑊. 𝑠𝑖𝑛 𝜃 (14)

𝐹𝑓 = 𝜇 . 𝑁 (15)

𝐹𝑓 = 𝜇. (𝑊. 𝑐𝑜𝑠 𝜃) (16)

𝐹𝑛𝑎𝑣 = 𝐹𝑔 + 𝐹𝑠 + 𝐹𝑝 − 𝐹𝑓 (17)

2.3. Fuzzy inference system

The 𝑘 and Ψ values as shown in (3) and (4) are gain values that are made adaptive based on the
fuzzy inference system [18], [19], in previous research, the robot detection area can be divided [7], [8]. Table
1 shows the fuzzy rule for static obstacles which deals with the division of the proxemics area (robot
detection area). The 𝑘 gain function is used to measure the extent to which the robot responds or reacts to the
influence of obstacles in this model. A higher value of gain k will result in a stronger response to the
obstacle, while a lower value of gain 𝑘 will result in a weaker response. In this paper, we also ignored the Ψ
value, since it does not have a significant impact and can be determined manually.

Table 1. Fuzzy rule for static obstacle

R Obstacle Distance
Intimate Personal Social Public
Obstacle Angle Front 50 25 25 25
Side-Front 25 25 12.5 12.5
Side 12.5 6.25 6.25 6.25

Figure 3 explains the division of the radius of the proxemic area [28], [29], delineating specific
zones based on interpersonal distances. These zones include intimate space (less than 45 cm), personal space
(45 to 120 cm), and social space (121 to 360 cm), while areas beyond these boundaries are classified as
public spaces. In the context of the SFM, the model primarily focuses on detecting and responding to forces
within the intimate, personal, and social spaces. Notably, public space, although not directly detected by the
SFM, is still factored into the decision-making process through the application of fuzzy logic. Fuzzy logic

IAES Int J Rob & Autom, Vol. 13, No. 2, June 2024: 131-139
IAES Int J Rob & Autom ISSN: 2722-2586  135

plays a crucial role in determining the adaptive gain value, enabling the robotic system to dynamically adjust
its behavior, considering the broader context of public spaces. This integration ensures a more nuanced and
adaptive response, accounting for the subtleties of social dynamics across different proxemic zones.

Figure 3. Proxemic area

Based on the partitioning of the proxemic area and the angle of approach by the robot [18], [19], a
membership function can be thoughtfully designed, as illustrated in Figure 4. Using this information, the
robot gains the ability to discern and react suitably to distinct proxemic zones, thereby optimizing its
behavior in alignment with human preferences and interaction constraints. This nuanced approach empowers
the robot to operate with heightened adaptability and responsiveness. By incorporating a tailored membership
function, the robot not only interprets its surroundings more effectively but also tailors its responses to
various proxemic zones, thereby enhancing its overall capacity to navigate and interact in a manner that is
attuned to human expectations and preferences.

Figure 4. Membership function

3. RESULTS AND DISCUSSION

The SFM simulation was conducted using the CoppeliaSim application using a differential drive
mobile robot (DDMR) 4 wheel. This implementation used various programming languages, including C++,
Lua, and Python, and connected them through the Robotic Operating System (ROS) communication system,
for full specification all hardware and software are in Table 2. The simulation environment used was set up
as shown in Table 2. We have developed Lua scripts in CoppeliaSim to control motors within the simulation,
enabling both linear and angular speed adjustments as depicted in Figure 5. The SFM algorithm, implemented
in C++, includes the calculation of forces on the inclined plane. The implementation of this algorithm is
distributed through the ROS communication system. Visualization of the simulation results is displayed using
OpenCV with the Python programming language, providing a clear visual representation of the robot’s
movement and its interaction with obstacles in the simulated environment. In the context of the robot, its
weight is set in the range of about 25 kg and I specified an inclined plane with a slope of approximately
15 degrees. All these settings are designed to achieve accurate and representative simulation results.

Robot navigation on inclined terrain using social force model (Muhammad Fariz Daffa)
136  ISSN: 2722-2586

Table 2. Simulation specifications

Hardware Description
Mini PC IntelNuc
Processor Intel(R) Core(TM) i5-10210U CPU @ 1.60 GHz
RAM 8192 MB DDR4
HDD 128 GB
Operating System and other Software Visual Studio Code

Figure 5. Defined Simulated environment in the CoppeliaSim

The robot will be made to move from the start position to the end position. On this journey, several
obstacles will be placed to test the robot’s navigation capabilities based on the SFM. Two types of SFM
models will be tested. First, the SFM without taking into account the force on the inclined plane is marked
with a yellow line, and second, the SFM that takes into account the force on the pre-made inclined plane with
a blue line, then the red dots are a representation of the distance reading from the lidar as can be seen in
Figure 6, while the corresponding travel times are depicted in Figure 7. The robot successfully navigates and
reaches the destination point by avoiding several obstacles using SFM.

Figure 6. Visualization of robot movement

(a) (b)

Figure 7. Time taken for the robot to reach the goal (a) without and (b) with force on an inclined plane

The test results reveal a significant difference in the time taken by the robot to reach the finish line.
When using the SFM model without considering the forces on the inclined plane, the required time duration

IAES Int J Rob & Autom, Vol. 13, No. 2, June 2024: 131-139
IAES Int J Rob & Autom ISSN: 2722-2586  137

is 63.506442 seconds according to Figure 7(a). However, when the robot takes into account the forces on the
inclined plane, it can be observed that the robot reaches the goal faster, approximately in 59.935089 seconds
according to Figure 7(b). The comparison of values before and after incorporating inclined force calculations
is visually presented in Figure 8. A detailed breakdown of these findings is provided in Table 3. Notably,
accounting for forces on the inclined plane enables the robot to reach its destination more swiftly by
identifying a more efficient route. In practical terms, this optimization reduces slipping and minimizes
wasteful movements, contributing to the overall enhanced performance of the robot.

Table 3. Result of the time taken and travel distance

Method SFM Without Force on an Inclined Plane SFM With Force on an Inclined Plane
Time 63.506442 59.935089
Distance 16.643172424591 16.095605425016

Figure 8. Changes in SFM forces against an inclined plane

In the graphical illustration in Figure 9, the force values in the SFM without considering the
inclination of the plane are represented by blue lines, while those that take into account the inclination of the
plane are shown in yellow. The force due to the inclination of the plane is represented in red, while the
change in pitch on the inclined plane is depicted in blue. It is important to note that when the robot passes
through an inclined section that is a descent (being in a downhill position), the system will calculate the force
arising from the inclination. As a result, the navigation force in the SFM will decrease, and therefore, the
speed of the robot on the inclined plane will also decrease in line with the change in navigation force. By
implementing this system, the speed of the robot when using the SFM becomes more adaptive when traveling
on an inclined plane. This results in robot navigation that is more stable and responsive to changing terrain
conditions, allowing the robot to move efficiently and safely even on surfaces with varying slopes.

4. CONCLUSION
In this work, we successfully overcome the main obstacle of the Social Force Model on inclined
terrain by developing an adaptive system that allows the robot to move efficiently and safely on inclined
terrain. The force adaptive control we implemented involves real-time detection of changes in surface slope,
enabling dynamic adjustment to frictional and parallel forces. Simulation results show a significant
improvement in the stability of the robot’s movement, reducing the risk of sudden changes in speed and
ensuring smoother navigation on inclined terrain. In this context, our research not only enriches the

Robot navigation on inclined terrain using social force model (Muhammad Fariz Daffa)
138  ISSN: 2722-2586

understanding of SFM applications in complex environments but also provides a foundation for the
development of future navigation technologies. The ability of robots to dynamically adapt their movements
to the terrain will be key in the development of robots that can operate reliably in various environmental
conditions. As such, this research not only has an important impact on academia but also opens up new
opportunities in the use of robotics technology in various practical applications, including the exploration of
difficult environments and hard-to-reach places. We are confident that the findings and approaches we
developed in this research will pave the way for further research and development of innovative and reliable
robot navigation solutions.

ACKNOWLEDGEMENTS
The authors would like to thank all members of Social Robotics and Smart System Application
(SRSSA) Research Group and the Center of Research and Community Services of Politeknik Elektronika
Negeri Surabaya (PENS) for their support, so we can complete this research. We also express our gratitude to
the anonymous reviewer for reviewing our paper.

REFERENCES
[1] A. F. B. Silva, A. V. Santos, M. A. Meggiolaro, and M. S. Neto, “A rough terrain traction control technique for all-wheel-drive
mobile robots,” Journal of the Brazilian Society of Mechanical Sciences and Engineering, vol. 32, no. 4, pp. 489–501, Dec. 2010,
doi: 10.1590/S1678-58782010000400011.
[2] C. Gehring et al., “Dynamic trotting on slopes for quadrupedal robots,” in IEEE International Conference on Intelligent Robots
and Systems, Sep. 2015, pp. 5129–5135. doi: 10.1109/IROS.2015.7354099.
[3] M. Focchi, A. del Prete, I. Havoutis, R. Featherstone, D. G. Caldwell, and C. Semini, “High-slope terrain locomotion for torque-
controlled quadruped robots,” Autonomous Robots, vol. 41, no. 1, pp. 259–272, May 2017, doi: 10.1007/s10514-016-9573-1.
[4] B. S. B. Dewantara and J. Miura, “Generation of a socially aware behavior of a guide robot using reinforcement learning,” in
2016 International Electronics Symposium (IES), Sep. 2016, pp. 105–110. doi: 10.1109/ELECSYM.2016.7860984.
[5] C. Wang, Y. Li, S. S. Ge, and T. H. Lee, “Adaptive control for robot navigation in human environments based on social force
model,” in 2016 IEEE International Conference on Robotics and Automation (ICRA), May 2016, pp. 5690–5695. doi:
10.1109/ICRA.2016.7487791.
[6] A. Colombo et al., “Behavioural templates improve robot motion planning with social force model in human environments,” in
2013 IEEE 18th Conference on Emerging Technologies and Factory Automation (ETFA), Sep. 2013, pp. 1–6. doi:
10.1109/ETFA.2013.6648081.
[7] B. S. B. Dewantara, “Building a socially acceptable navigation and behavior of a mobile robot using Q-learning,” in 2016
International Conference on Knowledge Creation and Intelligent Computing (KCIC), Nov. 2016, pp. 88–93. doi:
10.1109/KCIC.2016.7883630.
[8] G. Ferrer, A. Garrell, and A. Sanfeliu, “Robot companion: a social-force based approach with human awareness-navigation in
crowded environments,” in IEEE International Conference on Intelligent Robots and Systems, Nov. 2013, pp. 1688–1694. doi:
10.1109/IROS.2013.6696576.
[9] M. Luber, J. A. Stork, G. D. Tipaldi, and K. O. Arras, “People tracking with human motion predictions from social forces,” in
2010 IEEE International Conference on Robotics and Automation, May 2010, pp. 464–469. doi: 10.1109/ROBOT.2010.5509779.
[10] S. Silva, D. Paillacho, N. Verdezoto, and J. D. Hernandez, “Towards online socially acceptable robot navigation,” in IEEE
International Conference on Automation Science and Engineering, Aug. 2022, pp. 707–714. doi:
10.1109/CASE49997.2022.9926686.
[11] A. T. Rifqi, B. S. B. Dewantara, D. Pramadihanto, and B. S. Marta, “Fuzzy social force model for healthcare robot navigation and
obstacle avoidance,” in International Electronics Symposium 2021: Wireless Technologies and Intelligent Systems for Better
Human Lives, IES 2021 - Proceedings, Sep. 2021, pp. 445–450. doi: 10.1109/IES53407.2021.9594052.
[12] Ó. Gil, A. Garrell, and A. Sanfeliu, “Social robot navigation tasks: combining machine learning techniques and social force
model,” Sensors, vol. 21, no. 21, Oct. 2021, doi: 10.3390/s21217087.
[13] A. Garrell, L. Garza-Elizondo, M. Villamizar, F. Herrero, and A. Sanfeliu, “Aerial social force model: a new framework to
accompany people using autonomous flying robots,” in IEEE International Conference on Intelligent Robots and Systems, Sep.
2017, pp. 7011–7017. doi: 10.1109/IROS.2017.8206627.
[14] I. Sakour and H. Hu, “Robot-assisted crowd evacuation under emergency situations: A survey,” Robotics, vol. 6, no. 2, Apr. 2017,
doi: 10.3390/robotics6020008.
[15] B. Tang, C. Jiang, H. He, and Y. Guo, “Human mobility modeling for robot-assisted evacuation in complex indoor
environments,” IEEE Transactions on Human-Machine Systems, vol. 46, no. 5, pp. 694–707, Oct. 2016, doi:
10.1109/THMS.2016.2571269.
[16] J. Li, H. Zhang, and Z. Ni, “Improved social force model based on navigation points for crowd emergent evacuation,” Journal of
Information Processing Systems, vol. 16, no. 6, pp. 1309–1323, 2020, doi: 10.3745/JIPS.04.0199.
[17] X. Li, Y. Liang, M. Zhao, C. Wang, H. Bai, and Y. Jiang, “Simulation of evacuating crowd based on deep learning and social
force model,” IEEE Access, vol. 7, pp. 155361–155371, 2019, doi: 10.1109/ACCESS.2019.2949106.
[18] B. S. B. Dewantara and B. N. D. Ariyadi, “Adaptive behavior control for robot soccer navigation using fuzzy-based social force
model,” Smart Science, vol. 9, no. 1, pp. 14–29, Jan. 2021, doi: 10.1080/23080477.2021.1871799.
[19] A. Wibisana, B. S. B. Dewantara, and D. Pramadihanto, “Implementation of particle swarm optimization for optimizing fuzzy-
social force models in omnidirectional robot navigation systems (in Indonesia: Implementasi particle swarm optimization untuk
optimasi fuzzy-social force model pada sistem navigasi robot o,” Indonesian Journal of Computer Science, vol. 11, no. 2, Aug.
2022, doi: 10.33022/ijcs.v11i2.3076.
[20] X. T. Truong and T. D. Ngo, “Toward socially aware robot navigation in dynamic and crowded environments: a proactive social
motion model,” IEEE Transactions on Automation Science and Engineering, vol. 14, no. 4, pp. 1743–1760, Oct. 2017, doi:

IAES Int J Rob & Autom, Vol. 13, No. 2, June 2024: 131-139
IAES Int J Rob & Autom ISSN: 2722-2586  139

10.1109/TASE.2017.2731371.
[21] T. I. Lakoba, D. J. Kaup, and N. M. Finkelstein, “Modifications of the Helbing-Molnár-Farkas-Vicsek social force model for
pedestrian evolution,” Simulation, vol. 81, no. 5, pp. 339–352, May 2005, doi: 10.1177/0037549705052772.
[22] P. Regier, I. Shareef, and M. Bennewitz, “Improving navigation with the social force model by learning a neural network
controller in pedestrian crowds,” in 2019 European Conference on Mobile Robots (ECMR), Sep. 2019, pp. 1–6. doi:
10.1109/ECMR.2019.8870923.
[23] Z. Kausar, K. Stol, and N. Patel, “The effect of terrain inclination on performance and the stability region of two-wheeled mobile
robots,” International Journal of Advanced Robotic Systems, vol. 9, no. 5, Nov. 2012, doi: 10.5772/52894.
[24] A. V Pesterev and L. B. Rapoport, “Stabilization problem for a wheeled robot following a curvilinear path on uneven terrain,”
Journal of Computer and Systems Sciences International, vol. 49, no. 4, pp. 672–680, Aug. 2010, doi:
10.1134/S1064230710040155.
[25] X. Zhu, C. Qiu, and M. A. Minor, “Terrain inclination aided three-dimensional localization and mapping for an outdoor mobile
robot,” International Journal of Advanced Robotic Systems, vol. 10, no. 1, Jan. 2013, doi: 10.5772/54957.
[26] D. Helbing and P. Molnár, “Social force model for pedestrian dynamics,” Physical Review E, vol. 51, no. 5, pp. 4282–4286, May
1995, doi: 10.1103/PhysRevE.51.4282.
[27] J. Wang and M. Q. H. Meng, “Socially compliant path planning for robotic autonomous luggage trolley collection at airports,”
Sensors (Switzerland), vol. 19, no. 12, p. 2759, Jun. 2019, doi: 10.3390/s19122759.
[28] E. T. Hall, “Handbook for proxemic research,” in Handbook for Proxemic Research, Society for the Anthropology of Visual
Communication, 1974.
[29] P. Patompak, S. Jeong, I. Nilkhamhang, and N. Y. Chong, “Learning proxemics for personalized human–robot social interaction,”
International Journal of Social Robotics, vol. 12, no. 1, pp. 267–280, May 2020, doi: 10.1007/s12369-019-00560-9.

BIOGRAPHIES OF AUTHOR

Muhammad Fariz Daffa was born in Palembang, Indonesia, in 1999. He

received a bachelor’s degree in electrical engineering from the Polytechnic Institute of
Sriwijaya, and now he is pursuing a master’s degree in electrical engineering at the
Polytechnic Institute of Surabaya. His research interests are in robotics and kinematics. He can
be contacted via email at [email protected].

Bima Sena Bayu Dewantara is a lecturer in the Department of Informatics and

Computer Engineering from Politeknik Elektronika Negeri Surabaya, Indonesia. He received
the B.Eng. degree in Information Technology from Politeknik Elektronika Negeri Surabaya,
Indonesia, the M.Eng. degree in Electrical Engineering from Sepuluh Nopember Institute of
Technology, Indonesia, and the Dr. Eng. degree in Computer Science and Engineering from
Toyohashi University of Technology, Japan, in 2004, 2010, and 2016, respectively. Currently,
he works in the major field of autonomous intelligent systems which covers pattern
recognition, computer vision, machine learning, signal processing, robotics system, and
human-machine interaction. He has published more than 35 international dan national journal
papers and more than 75 international dan national paper proceedings. He received The Best
Achievement Award from Toyohashi University of Technology, Japan, in 2016 and also
received some Best Paper Awards in 2016, 2018, 2019, 2020, 2021, and 2022. He can be
contacted at [email protected].

Setiawardhana is a lecturer in the Department of Computer Engineering from

Politeknik Elektronika Negeri Surabaya, Indonesia. He received the bachelor’s, master’s and
doctoral degrees in electronics engineering from Institut Teknologi Sepuluh Nopember,
Indonesia, in 2000, 2010, and 2021. His research interests include the areas of artificial
intelligence, image processing, internet of things, and robotics. He can be contacted at
[email protected].

Robot navigation on inclined terrain using social force model (Muhammad Fariz Daffa)