CineWild: Balancing Art and Robotics for Ethical Wildlife Documentary Filmmaking
Organizations: Service Robotics Lab, Universidad Pablo de Olavide, Seville, Spain.
Abstract
Unmanned aerial vehicles (UAVs) have become powerful tools across domains from industry to the arts. In documentary filmmaking, they offer dynamic, otherwise unreachable perspectives, transforming how stories are told. Wildlife documentaries especially benefit, yet UAVs also raise ethical concerns like the risk of disturbing the animals they aim to capture. This paper introduces CineWild, a model predictive control (MPC) formulation for autonomous aerial cinematography that combines robotics, cinematography, and ethics. Building on an MPC solution that jointly controls drone motion and camera focal length, CineWild dynamically adjusts flight paths and camera settings to balance cinematic quality with animal welfare. Key features include adaptive zoom for filming from acoustic and visual safe distances, path-planning that avoids an animal's field of view, and smooth, low-noise maneuvers. CineWild exemplifies interdisciplinary innovation, bridging engineering, visual storytelling, and environmental ethics. We evaluate the formulation in simulation, with ablation studies and a comparison against a cinematographic baseline and a welfare-aware baseline under identical weights, and we demonstrate the perception and camera-actuation pipeline in real-world proof-of-concept trials. We will release the code.
Figures & tables
| Wildlife | Welfare terms | Zoom control | Subject FoV | Proximity | |
|---|---|---|---|---|---|
| Nägeli et al. [ 22 ] | – | – | – | – | safety only |
| Bonatti et al. [ 13 ] | – | – | – | – | safety only |
| Krátký et al. [ 1 ] | – | – | – | – | safety only |
| CineMPC [ 7 ] | – | – | ✓ | – | safety only |
| CineWild (ours) | ✓ | ✓ | ✓ | ✓ | ✓ |
| Exp | ||||||
|---|---|---|---|---|---|---|
| E1 | 15 | 1 | 2 | 1 | 0 | 250 |
| E2 | 15 | 1 | 2 | 0.5 | 10 | 100 |
| E1 (giraffe) | 5 m | 20 m | 12 m | |||
| E2 (tiger) | 5 m | 8 m | 12 m | |||
| CineWild | CineMPC | CineMPC + bubble | |
| min / mean [m] | 14.6 / 23.5 | 6.6 / 10.9 | 14.5 / 19.8 |
| time [%] | 15.0 | 100 | 22.2 |
| time within 1 m of [%] | 9.7 | – | 90.9 |
| / [px] | 17.0 / 14.1 | 17.9 / 12.3 | 40.3 / 26.3 |
| [deg] | 27.2 | 10.9 | 26.8 |
| focal length [mm] | 71 | 35 | 55 |
| CineWild | CineMPC | CineMPC + bubble | |
| time in animal FoV [%] | 2.5 | 73.1 | 74.7 |
| drone position in animal image [px] | 180 | 310 | 314 |
| min [m] | 9.1 | 8.8 | 9.1 |
| / [px] | 39.3 / 9.6 | 10.4 / 9.2 | 8.7 / 9.5 |
| [deg] | 14.6 | 1.2 | 1.4 |
| focal length [mm] | 49 | 50 | 49 |
| Config | min [m] | [%] | [%] | [px] | [px] | [deg] | [m/s 2 ] | |
|---|---|---|---|---|---|---|---|---|
| E1 | CineWild | 14.6 | 15.0 | 0 | 17.0 | 14.1 | 27.2 | 0.11 |
| w/o | 6.5 | 100 | 0 | 21.2 | 13.9 | 15.0 | 0.16 | |
| w/o FoV | 14.5 | 11.9 | 0 | 15.5 | 13.0 | 28.5 | 0.10 | |
| w/o | 14.6 | 11.7 | 0 | 13.5 | 15.1 | 26.7 | 0.13 | |
| E2 | CineWild | 9.1 | 0 | 2.5 | 39.3 | 9.6 | 14.6 | 0.19 |
| w/o | 8.6 | 1.5 | 3.2 | 55.4 | 12.7 | 14.8 | 0.20 |
| rear view | [deg] | / [px] | [m/s 2 ] | [%] | |
|---|---|---|---|---|---|
| 5 (2 s) | 10/10 | 27.2 | 17.0 / 14.1 | 0.11 | 15.0 |
| 10 (4 s) | 5/5 | 45.6 | 26.7 / 16.2 | 0.10 | 13.3 |
| Config | runs | [%] | [m] | / [px] | [m/s 2 ] | |
|---|---|---|---|---|---|---|
| reference | , m | 10 | 22.5 | 24.0 | 13.7 / 15.8 | 0.10 |
| 5 | 5 | 23.8 | 23.8 | 14.2 / 14.8 | 0.09 | |
| 45 | 5 | 22.0 | 25.5 | 13.7 / 15.1 | 0.10 | |
| 10 m | 5 | 3.8 | 16.3 | 21.4 / 11.9 | 0.11 | |
| 30 m | 5 | 48.9 | 27.2 | 18.3 / 16.3 | 0.11 |
Aerial Robotics
Aerial robotics studies the design, perception, control, and coordination of flying robots for navigation, inspection, transport, and physical interaction.
See all 354 papers