DRAMA - DRiver Activity MApping
| Reference number | |
| Coordinator | RISE Research Institutes of Sweden AB - RISE Viktoria, Göteborg |
| Funding from Vinnova | SEK 4 850 000 |
| Project duration | March 2018 - February 2020 |
| Status | Completed |
| Venture | Electronics, software and communication - FFI |
| End-of-project report | 2017-05503engelska.pdf (pdf, 6606 kB) |
Purpose and goal
DRAMA project studies possible mappings and modelings of people´s behaviour in vehicles with computer vision and machine learning to enable improved security and UX functionality. The project develops knowledge, frameworks and prototypes. Including: - List of driver / passenger activities in cabin that are important for safety and comfort. - Mappings of body posture, facial expressions, main position, objects and movements - Modeling of in-cabin activities / interactions. - Potential Applications
Expected results and effects
The project delivers the following results: - Lists of in-cabin activity scenarios that are related to safety/comfort in SAE2-5 vehicles. - Mappings and modellings of in-cabin activities and recognition modules providing important information from different perspectives for UX/safety applications in highly automated vehicles. This will help to correctly adjust vehicle behaviours based on in-cabin context for safer and better comfort rides. Understanding of what peoples are doing inside the cabin can also be used for prediction of their intentions and expectations.
Planned approach and implementation
The project is a collaboration in research and development. The works were carried out in an iterative approach involving the business areas: brainstorming workshops, interviews, data analysis, and prototype developments. Stake holders from a research institute, SME and advisory board from OEMs were involved, forming a strong and flexible team. The continuous supports and high enthusiasm from all stakeholders to keep the project towards its main goals while conducting different attempts are the key success factors.