Ethical Issues

Problem

The drone system needs to perform better than non-UAV rescue teams across a variety of disasters. If the system fails to accomplish a task, this poses an ethical dilemma in the context of convincing organizations to purchase systems that do not live up to the promise of saving more lives.

Solution

Human oversight is critical to the long-term use of this technology. Total reliance should not be a goal due to the grave consequences of mistakes. A potential solution is for a team to review the decisions made by the system post-operation to concisely evaluate the performance in a variety of environments and provide corrections where needed. 

Problem

A path algorithm that searches probabilistically raises questions about how to fairly search a disaster area.

Solution 

Consult with rescue organizations on their process, and moral stances on search and rescue to ensure our search algorithm is ethically consistent with the current norms of search and rescue.

Consider creating a hybrid approach where the search algorithm prioritizes areas with a high statistical probability of casualties while also delegating resources to search less likely areas.

Problem

Potential for misidentification of people based on bias in data.

Solution:

Collect a wide range of sample data from various demographics. You must be conscious of training data and the bias it may hold and be able to fill in demographic data when needed.