A company must collect 6 TB of proprietary-format experiment data from sensors in a remote location with no internet. Sensors can periodically upload files via FTP, but they cannot act as FTP servers or use other protocols. The company needs to centralize the uploads locally and move the data to AWS object storage as soon as possible after the experiment. Which solution satisfies these constraints?

The correct solution leverages AWS Snowcone's compute capabilities to host an FTP server. Sensors can only upload via FTP and cannot act as servers, so a local FTP server is essential. Snowcone's Amazon Linux 2 AMI allows running an EC2 instance to install and configure this server. After data collection, the Snowcone device is returned to AWS for direct upload to Amazon S3, meeting the requirement for moving data to object storage. The first incorrect option, Snowball Edge with DataSync and NFS, fails because sensors only support FTP, not NFS, and cannot be configured as DataSync sources. The second incorrect option, Snowcone with an EC2 instance to download data, is incorrect because sensors upload, not allow downloads. The fourth incorrect option, Snowcone with FSx and DataSync, is incorrect because sensors only support FTP, not direct uploads to FSx or DataSync, and the data is destined for S3, not EBS.