Would mass-produced closed loop be a viable solution here? I'd imagine the engineering costs incurred ahead of time would add to quite a bit, but I'm unqualified to make any interesting judgment on any final balance of costs.