Autonomous Recognition of Collective Behaviour in Robot Swarms

Kasmarik, Kathryn; Abpeikar, Shadi; Khan, Md. Mohiuddin; Khattab, Noha; Barlow, Michael; Garratt, Matt

doi:10.1007/978-3-030-64984-5_22

Cited by 5 publications

(6 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The simulation on Pioneer 3DX was selected since the scenarios were available from Kasmarik et al ( 2020 ). As an example of real robots the Sphero BOLT robots are selected, as they permit us to run the short duration swarming experiments in this paper without need for an external positioning system.…”

Section: Experiments and Discussionmentioning

confidence: 99%

“…To exploit the human ability to recognize collective behavior recognition, one recent approach collected data from humans regarding their opinion of different behaviors. Then this labeled data was used to train a machine to mimic human recognition of swarming collective motion (Kasmarik et al, 2020 ). Human labels were collected via an online survey.…”

Section: Methodology: Collective Motion Tuning In Robots Using Transf...mentioning

confidence: 99%

“…Various machine learning algorithms were tried to see which could best mimic human recognition using this dataset as a training set. Following these experiments, the decision tree resulted in the maximum accuracy (Kasmarik et al, 2020 ). A knowledge base of 73 if-then rules was extracted from a pruned version of this decision tree (Abpeikar et al, 2022a ).…”

Section: Methodology: Collective Motion Tuning In Robots Using Transf...mentioning

confidence: 99%

“…It is hard for a machine, however, humans can recognize this behavior easily. Hence, recent works used human perception to train machines for automatic collective behavior recognition in simulation (Kasmarik et al, 2020 ; Abpeikar et al, 2022b ). Following this approach, some success has been achieved for automatic collective motion generation with RL.…”

Section: Background and Related Workmentioning

confidence: 99%

“…More details on the different simulated scenarios are provided in Section 4.1. For more information on Pioneer 3DX robots, see e.g., Kasmarik et al ( 2020 ).…”

Section: Background and Related Workmentioning

confidence: 99%

See 4 more Smart Citations

Iterative transfer learning for automatic collective motion tuning on multiple robot platforms

2023

Self Cite

View full text Add to dashboard Cite

This paper proposes an iterative transfer learning approach to achieve swarming collective motion in groups of mobile robots. By applying transfer learning, a deep learner capable of recognizing swarming collective motion can use its knowledge to tune stable collective motion behaviors across multiple robot platforms. The transfer learner requires only a small set of initial training data from each robot platform, and this data can be collected from random movements. The transfer learner then progressively updates its own knowledge base with an iterative approach. This transfer learning eliminates the cost of extensive training data collection and the risk of trial-and-error learning on robot hardware. We test this approach on two robot platforms: simulated Pioneer 3DX robots and real Sphero BOLT robots. The transfer learning approach enables both platforms to automatically tune stable collective behaviors. Using the knowledge-base library the tuning procedure is fast and accurate. We demonstrate that these tuned behaviors can be used for typical multi-robot tasks such as coverage, even though they are not specifically designed for coverage tasks.

show abstract

Section: Experiments and Discussionmentioning

confidence: 99%

Section: Methodology: Collective Motion Tuning In Robots Using Transf...mentioning

confidence: 99%

Section: Methodology: Collective Motion Tuning In Robots Using Transf...mentioning

confidence: 99%

Section: Background and Related Workmentioning

confidence: 99%

“…More details on the different simulated scenarios are provided in Section 4.1. For more information on Pioneer 3DX robots, see e.g., Kasmarik et al ( 2020 ).…”

Section: Background and Related Workmentioning

confidence: 99%

See 3 more Smart Citations