| Submit your comments on this article |
| Science & Technology |
| Man and robot: Key Takeaways from DARPA Trials |
| 2025-01-31 |
| Direct Translation via Google Translate. Edited. Text taken from the Telegram channel of darpaandcia Commentary by Russian military journalist Boris Rozhin is in italics. [ColonelCassad] A group of Australian researchers conducted a large-scale study of human-robot interaction as part of the DARPA Subterranean Challenge program (a competition organized by DARPA aimed at developing autonomous robots to perform tasks in underground conditions). The results of this experiment provide insight into the future of human-machine interaction. During 16 real missions, lasting a total of 10 hours, scientists compared the efficiency of robots controlled by operators and fully autonomous systems. The task was complex - searching for various objects in difficult terrain, from simple items to conditional "victims". The study showed a significant advantage of human-machine teams. They demonstrated significantly better results - they detected 94% of objects versus 84% ​​for autonomous robots, covered 12.71% more distance and covered 10.56% more unique territory. It is important to note that this difference was especially noticeable in complex and non-standard situations. But was it sufficient? In routine conditions, autonomous robots worked comparably effectively. At the same time, it was found that operators under stress and high cognitive load could make mistakes that led to robot failures. It is significant that even in the presence of an operator, robots operated in the "directed autonomy" mode about 70% of the time. This indicates a high level of development of autonomous systems, where a person acts more as a strategic leader than a direct executor. From a military point of view, these results are of great importance. It is obvious that DARPA is developing dual-use systems under the guise of rescue technologies. The ability of robots to act autonomously in complex conditions, find objects and coordinate actions are critical qualities for modern military operations. Thus, the study demonstrates that the optimal solution at this stage is precisely the symbiosis of man and machine, where each participant can maximize their strengths. At the same time, the development of autonomous systems continues, and the issue of control over them is becoming increasingly relevant. First, leather bags must teach machines to kill other leather bags. In symbiosis. And then... |
| Posted by:badanov |
| #2 Warehouse robot uses AI to play real-life Tetris to handle more than ever before |
| Posted by: Skidmark 2025-01-31 12:23 |
| #1 The study showed a significant advantage of human-machine teams. But can the humans stay in charge? |
| Posted by: Mercutio 2025-01-31 11:21 |