Skip to content

Comparing Model-Based and Data-Driven Controllers for an Autonomous Vehicle Task

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Standard

Comparing Model-Based and Data-Driven Controllers for an Autonomous Vehicle Task. / López Pulgarín, Erwin José; Irmak, Tugrul; Variath Paul, Joel; Meekul, Arisara; Herrmann, Guido; Leonards, Ute.

Towards Autonomous Robotic Systems: 19th Annual Conference, TAROS 2018, Bristol, UK July 25-27, 2018, Proceedings. Springer, Cham, 2018. p. 170-182 (Lecture Notes in Artificial Intelligence; Vol. 10965).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Harvard

López Pulgarín, EJ, Irmak, T, Variath Paul, J, Meekul, A, Herrmann, G & Leonards, U 2018, Comparing Model-Based and Data-Driven Controllers for an Autonomous Vehicle Task. in Towards Autonomous Robotic Systems: 19th Annual Conference, TAROS 2018, Bristol, UK July 25-27, 2018, Proceedings. Lecture Notes in Artificial Intelligence, vol. 10965, Springer, Cham, pp. 170-182. https://doi.org/10.1007/978-3-319-96728-8_15

APA

López Pulgarín, E. J., Irmak, T., Variath Paul, J., Meekul, A., Herrmann, G., & Leonards, U. (2018). Comparing Model-Based and Data-Driven Controllers for an Autonomous Vehicle Task. In Towards Autonomous Robotic Systems: 19th Annual Conference, TAROS 2018, Bristol, UK July 25-27, 2018, Proceedings (pp. 170-182). (Lecture Notes in Artificial Intelligence; Vol. 10965). Springer, Cham. https://doi.org/10.1007/978-3-319-96728-8_15

Vancouver

López Pulgarín EJ, Irmak T, Variath Paul J, Meekul A, Herrmann G, Leonards U. Comparing Model-Based and Data-Driven Controllers for an Autonomous Vehicle Task. In Towards Autonomous Robotic Systems: 19th Annual Conference, TAROS 2018, Bristol, UK July 25-27, 2018, Proceedings. Springer, Cham. 2018. p. 170-182. (Lecture Notes in Artificial Intelligence). https://doi.org/10.1007/978-3-319-96728-8_15

Author

López Pulgarín, Erwin José ; Irmak, Tugrul ; Variath Paul, Joel ; Meekul, Arisara ; Herrmann, Guido ; Leonards, Ute. / Comparing Model-Based and Data-Driven Controllers for an Autonomous Vehicle Task. Towards Autonomous Robotic Systems: 19th Annual Conference, TAROS 2018, Bristol, UK July 25-27, 2018, Proceedings. Springer, Cham, 2018. pp. 170-182 (Lecture Notes in Artificial Intelligence).

Bibtex

@inproceedings{6b74c54b2d8148a18e1ee2794d1cbb7d,
title = "Comparing Model-Based and Data-Driven Controllers for an Autonomous Vehicle Task",
abstract = "The advent of autonomous vehicles comes with many questions from an ethical and technological point of view. The need for high performing controllers, which show transparency and predictability is crucial to generate trust in such systems. Popular data-driven, black box-like approaches such as deep learning and reinforcement learning are used more and more in robotics due to their ability to process large amounts of information, with outstanding performance, but raising concerns about their transparency and predictability. Model-based control approaches are still a reliable and predictable alternative, used extensively in industry but with restrictions of their own. Which of these approaches is preferable is difficult to assess as they are rarely directly compared with each other for the same task, especially for autonomous vehicles. Here we compare two popular approaches for control synthesis, model-based control i.e. Model Predictive Controller (MPC), and data-driven control i.e. Reinforcement Learning (RL) for a lane keeping task with speed limit for an autonomous vehicle; controllers were to take control after a human driver had departed lanes or gone above the speed limit. We report the differences between both control approaches from analysis, architecture, synthesis, tuning and deployment and compare performance, taking overall benefits and difficulties of each control approach into account.",
author = "{L{\'o}pez Pulgar{\'i}n}, {Erwin Jos{\'e}} and Tugrul Irmak and {Variath Paul}, Joel and Arisara Meekul and Guido Herrmann and Ute Leonards",
year = "2018",
month = "7",
day = "21",
doi = "10.1007/978-3-319-96728-8_15",
language = "English",
isbn = "9783319967271",
series = "Lecture Notes in Artificial Intelligence",
publisher = "Springer, Cham",
pages = "170--182",
booktitle = "Towards Autonomous Robotic Systems",
address = "Switzerland",

}

RIS - suitable for import to EndNote

TY - GEN

T1 - Comparing Model-Based and Data-Driven Controllers for an Autonomous Vehicle Task

AU - López Pulgarín, Erwin José

AU - Irmak, Tugrul

AU - Variath Paul, Joel

AU - Meekul, Arisara

AU - Herrmann, Guido

AU - Leonards, Ute

PY - 2018/7/21

Y1 - 2018/7/21

N2 - The advent of autonomous vehicles comes with many questions from an ethical and technological point of view. The need for high performing controllers, which show transparency and predictability is crucial to generate trust in such systems. Popular data-driven, black box-like approaches such as deep learning and reinforcement learning are used more and more in robotics due to their ability to process large amounts of information, with outstanding performance, but raising concerns about their transparency and predictability. Model-based control approaches are still a reliable and predictable alternative, used extensively in industry but with restrictions of their own. Which of these approaches is preferable is difficult to assess as they are rarely directly compared with each other for the same task, especially for autonomous vehicles. Here we compare two popular approaches for control synthesis, model-based control i.e. Model Predictive Controller (MPC), and data-driven control i.e. Reinforcement Learning (RL) for a lane keeping task with speed limit for an autonomous vehicle; controllers were to take control after a human driver had departed lanes or gone above the speed limit. We report the differences between both control approaches from analysis, architecture, synthesis, tuning and deployment and compare performance, taking overall benefits and difficulties of each control approach into account.

AB - The advent of autonomous vehicles comes with many questions from an ethical and technological point of view. The need for high performing controllers, which show transparency and predictability is crucial to generate trust in such systems. Popular data-driven, black box-like approaches such as deep learning and reinforcement learning are used more and more in robotics due to their ability to process large amounts of information, with outstanding performance, but raising concerns about their transparency and predictability. Model-based control approaches are still a reliable and predictable alternative, used extensively in industry but with restrictions of their own. Which of these approaches is preferable is difficult to assess as they are rarely directly compared with each other for the same task, especially for autonomous vehicles. Here we compare two popular approaches for control synthesis, model-based control i.e. Model Predictive Controller (MPC), and data-driven control i.e. Reinforcement Learning (RL) for a lane keeping task with speed limit for an autonomous vehicle; controllers were to take control after a human driver had departed lanes or gone above the speed limit. We report the differences between both control approaches from analysis, architecture, synthesis, tuning and deployment and compare performance, taking overall benefits and difficulties of each control approach into account.

U2 - 10.1007/978-3-319-96728-8_15

DO - 10.1007/978-3-319-96728-8_15

M3 - Conference contribution

SN - 9783319967271

T3 - Lecture Notes in Artificial Intelligence

SP - 170

EP - 182

BT - Towards Autonomous Robotic Systems

PB - Springer, Cham

ER -