Project ATTIS

Project ATTIS (Aerial Terrain & Tracking Intelligence System) envisions developing a long-range autonomous with imaging sensors for conservation efforts in protected natural reserves.

Illegal farming, mining, and poaching have been great problems in such nature areas. The conventional method of surveillance is done through ground patrols that visit these areas one by one. This can be considered ineffective due to limited staff and tough access to the protected areas. Ground patrols are not very efficient in distributing resources because the areas are too large for patrols to find issues in time. As a result, resources from international NGOs, such as the WWF, are not being utilized optimally. Project ATTIS aims to provide a more efficient substitution for this problem.

For example, in Chizarira National Park in Zimbabwe NGOs pay for 4-6 flights of a conventional aircraft per year. This amount of flights is not enough for consistent surveillance. There are other UAV options available on the market but they are very expensive. Some examples are SB4 Phoenix ($25,000), Atmos UAV (€17,000), and BAE PHASA 35. A more affordable, long-range platform for imaging sensors that can fly autonomously over waypoints would provide a way for data to be gathered more efficiently, with lower costs and with less environmental impact.

Project ATTIS aims to help conservation efforts in the Chizarira National Park. [Picture by ‘Dissoxciate’ via Wikimedia Commons, CC BY-SA 4.0, unaltered]

Possible nature reservation initiatives that can be significantly aided by Project ATTIS include Chizarira National Park and Charisa Safari Area Protection, Rhino Reintroductions, Communication Enhancement, and Infrastructure inspection. Further applications may comprise early fire detection systems, and search & rescue.

Current Status/Scope

The project is in its early development phase. The short-term goal is to successfully complete an autonomous way-point experiment.

Currently, the preliminary sizing of the UAV is in progress. A range of payload instruments (including infra-red, quality stabilized cameras) are considered, and reinforcement learning implementations are researched.