Autonomous Cargo Bike Fleets – Approaches for AI-Based Trajectory Forecasts of Road Users
Automated cargo bikes are intended to complement public transportation in a sharing concept and provide an alternative transportation option for people and goods. In highly automated driving without a seated user, real-time trajectory prediction of other road users is crucial for collision avoidance...
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Sciendo
2023-02-01
|
| Series: | Transport and Telecommunication |
| Subjects: | |
| Online Access: | https://doi.org/10.2478/ttj-2023-0006 |
| _version_ | 1797848525852114944 |
|---|---|
| author | Sass Stefan Höfer Markus Schmidt Michael Schmidt Stephan |
| author_facet | Sass Stefan Höfer Markus Schmidt Michael Schmidt Stephan |
| author_sort | Sass Stefan |
| collection | DOAJ |
| description | Automated cargo bikes are intended to complement public transportation in a sharing concept and provide an alternative transportation option for people and goods. In highly automated driving without a seated user, real-time trajectory prediction of other road users is crucial for collision avoidance with other motor vehicles or vulnerable road users (VRU). For this purpose, moving obstacles are detected by environmental sensors and classified and tracked using object detection and tracking algorithms. The current and past position data as well as environmental information are used to predict future positions. In this paper, we present several AI-based trajectory prediction models that are specifically suited for this use case. Our focus is not only on the accuracy of trajectory prediction, but additionally on a robust, real-time and practical application. We consider models that can predict the trajectories with position estimation or distributions for position estimation for each time step in the future. For this aim, we present generative network structures based on Conditional Variational Autoencoder (CVAE) in different variants. After training, the models are integrated into our production system and their computation time is determined on the hardware we use. |
| first_indexed | 2024-04-09T18:28:59Z |
| format | Article |
| id | doaj.art-b2d10543eaa4420bad4375d71b7d0609 |
| institution | Directory Open Access Journal |
| issn | 1407-6179 |
| language | English |
| last_indexed | 2024-04-09T18:28:59Z |
| publishDate | 2023-02-01 |
| publisher | Sciendo |
| record_format | Article |
| series | Transport and Telecommunication |
| spelling | doaj.art-b2d10543eaa4420bad4375d71b7d06092023-04-11T17:28:19ZengSciendoTransport and Telecommunication1407-61792023-02-01241556410.2478/ttj-2023-0006Autonomous Cargo Bike Fleets – Approaches for AI-Based Trajectory Forecasts of Road UsersSass Stefan0Höfer Markus1Schmidt Michael2Schmidt Stephan31Otto-von-Guericke Universität MagdeburgUniversitätsplatz 2, 39106Magdeburg, Germany1Otto-von-Guericke Universität MagdeburgUniversitätsplatz 2, 39106Magdeburg, Germany1Otto-von-Guericke Universität MagdeburgUniversitätsplatz 2, 39106Magdeburg, Germany1Otto-von-Guericke Universität MagdeburgUniversitätsplatz 2, 39106Magdeburg, GermanyAutomated cargo bikes are intended to complement public transportation in a sharing concept and provide an alternative transportation option for people and goods. In highly automated driving without a seated user, real-time trajectory prediction of other road users is crucial for collision avoidance with other motor vehicles or vulnerable road users (VRU). For this purpose, moving obstacles are detected by environmental sensors and classified and tracked using object detection and tracking algorithms. The current and past position data as well as environmental information are used to predict future positions. In this paper, we present several AI-based trajectory prediction models that are specifically suited for this use case. Our focus is not only on the accuracy of trajectory prediction, but additionally on a robust, real-time and practical application. We consider models that can predict the trajectories with position estimation or distributions for position estimation for each time step in the future. For this aim, we present generative network structures based on Conditional Variational Autoencoder (CVAE) in different variants. After training, the models are integrated into our production system and their computation time is determined on the hardware we use.https://doi.org/10.2478/ttj-2023-0006autonomous drivingcargo bikeurban mobilitytrajectory predictionneural networksdeep learning |
| spellingShingle | Sass Stefan Höfer Markus Schmidt Michael Schmidt Stephan Autonomous Cargo Bike Fleets – Approaches for AI-Based Trajectory Forecasts of Road Users Transport and Telecommunication autonomous driving cargo bike urban mobility trajectory prediction neural networks deep learning |
| title | Autonomous Cargo Bike Fleets – Approaches for AI-Based Trajectory Forecasts of Road Users |
| title_full | Autonomous Cargo Bike Fleets – Approaches for AI-Based Trajectory Forecasts of Road Users |
| title_fullStr | Autonomous Cargo Bike Fleets – Approaches for AI-Based Trajectory Forecasts of Road Users |
| title_full_unstemmed | Autonomous Cargo Bike Fleets – Approaches for AI-Based Trajectory Forecasts of Road Users |
| title_short | Autonomous Cargo Bike Fleets – Approaches for AI-Based Trajectory Forecasts of Road Users |
| title_sort | autonomous cargo bike fleets approaches for ai based trajectory forecasts of road users |
| topic | autonomous driving cargo bike urban mobility trajectory prediction neural networks deep learning |
| url | https://doi.org/10.2478/ttj-2023-0006 |
| work_keys_str_mv | AT sassstefan autonomouscargobikefleetsapproachesforaibasedtrajectoryforecastsofroadusers AT hofermarkus autonomouscargobikefleetsapproachesforaibasedtrajectoryforecastsofroadusers AT schmidtmichael autonomouscargobikefleetsapproachesforaibasedtrajectoryforecastsofroadusers AT schmidtstephan autonomouscargobikefleetsapproachesforaibasedtrajectoryforecastsofroadusers |