The Never-Sleeping Crewman: Costs, Consumption, and Nanobots (AI for Sailing Boats)
29 de março de 2026 · 36 min · 4K visualizações
Man reads AI emergency email template on tablet in cabin
O que este vídeo aborda
Max, aboard the Beneteau Oceanis 50 Y2K in Australia, installs and demonstrates an AI agent called Nanobot as a permanent onboard crew member, interfacing with it via Telegram under the name Enea.▶ 7:52 Nanobot runs on 1 GB of RAM, written in 4,000 lines of Python — a fraction of the 400,000-line OpenCLO framework Max first attempted and abandoned because it required at least 8 GB and took days to configure.▶ 6:28 Nanobot was operational in under an hour.▶ 6:48 Max connects it to SignalK for live wind data, a Cerbo GX for battery monitoring (SOC, voltage, current, amp-hours), and solar production figures; Enea reads 597 W / 44 A solar output and 2.2 kWh charged that morning.▶ 14:49 Critically, Enea can remotely switch the battery charger on and off through the Cerbo, demonstrated live.▶ 15:20 It also pulls weather forecasts from Open Meteo, retrieves satellite imagery of the boat's position, and stores boat documents — including the Pantenius insurance certificate — sending them as Telegram attachments on demand.▶ 17:53 Max demonstrates a simulated collision claim: Enea drafts a complete incident email template to Pantenius with emergency contact numbers drawn from the policy document.▶ 20:57 Max notes he has not yet connected Enea to his email or calendar due to security concerns.▶ 17:56 A lightweight local AI agent running on minimal hardware can monitor onboard systems, control equipment remotely, and handle boat administration through a standard messaging app.
Barcos neste vídeo
Mais de Ale & Max - Y2K Sailing Adventure
De Ale & Max - Y2K Sailing Adventure — veja o barco, a tripulação e cada vídeo analisado do canal.