Mythos AI and lomarlabs deploy sea-pilot AI assistance
David Thomas
September 15, 2025
Illustrative image of tanker for article on AI sea navigation aids.


American maritime technology company Mythos AI has completed the installation of its Advanced Pilot Assistance System (APAS) aboard the vessel CB Pacific, a chemical cargo ship owned by CB Tankers.

The APAS project is designed to manage navigation at sea. Unlike most traditional systems in this space that depend largely on machine vision, APAS uses a radar-first approach, in combination with other sensing technologies. It connects directly to a ship’s radar and streamlines data so it can be used to alert a ship’s human crew if necessary during a journey. APAS is designed to support mariners, reduce cognitive load, and improve situational awareness, yet keep human judgement at the centre of navigation.

Geoff Douglass, CEO of Mythos AI, said, “Our goal isn’t to replace the crew. It’s to equip them with next-generation capabilities. By integrating our proprietary radar perception, machine vision, and intelligent alerting with the vessel’s dynamics, APAS transforms complex situations into clear, actionable decisions, enhancing safety and operational resilience.”

The CB Pacific was chosen for testing due to its predictable routes and reliable Furuno radar. The trial follows the first installation of APAS on a Southern Devall towboat on the Mississippi River in August 2025, and will be a year-long experiment aiming to introduce next-gen bridge intelligence to commercial shipping.

“Partnering with lomarlabs and CB Tankers enables APAS to capture and retain the expertise of master mariners and the navigational norms of ports worldwide […] We are validating performance at scale and laying the groundwork for broader fleet-wide adoption,” Douglass said.

The APAS system’s year-long trial will help ensure it functions safely and effectively in real-world conditions, and will check it can follow international regulations compliant with the COLREG (Convention on the International Regulations for Preventing Collisions at Sea).

CB Tankers is part of the Lomar group of companies. Managing director of lomarlabs, Stylianos Papageorgiou, said progress in maritime AI only comes from operational testing. “Real innovation doesn’t happen in pitch decks. It happens in real-time operations, port calls, dry docks, and sea trials.”

With increased interest from the defence sector, the APAS project represents a step toward adoption of AI-driven navigation systems in commercial and strategic maritime operations.

(Image source: “The M/T Carry on the Baltic Sea” by Mustang Joe is marked with CC0 1.0.)



Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and co-located with other leading technology events. Click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

https://youtu.be/BBpyQjhG51A?si=YKbB5LRROB2g5P2k

#MGBOX
#Google
#Google的創業史！

全球市場研究機構 TrendForce 針對 2025 年科技產業發展，整理十大重點趨勢。


生成式 AI 引領革新：人形化與服務型機器人迎來全新升級
隨著AI與機械動力技術日趨成熟，加諸NVIDIA、Tesla等大廠積極布局，機器人議題2025年將持續受市場關注。就技術發展而言，軟體平台著眼機器學習訓練與數位孿生模擬，整機型態則聚焦協作機器人、移動式機械手臂與人型機器人，以適應各式環境與人機協作互動。

其中，人型機器人隨美、中廠商積極投入，2025年起將逐步實現量產，預估2024年至2027年全球人型機器人市場規模之年複合成長率將達154%、產值有望一舉突破20億美元。倘看整體應用場域，相較於工業型仍以手臂撿貨為主，服務型機器人藉由生成式AI可支援多模態交流互動、檢索資訊、摘要文本、擬定排程等場景，帶出機動性高、陪伴性強、功能面廣等效益，將成為來年機器人發展重心。

技術革新推動市場標配：AI 筆電在 2025 年滲透率將達 21.7%
隨技術迅速發展，具有AI功能的筆記型電腦未來幾年內將逐漸成為市場標配。預計2025年AI筆電的滲透率將達到21.7%，並在2029年攀升至接近80%。而AI筆電的增量也將成為Arm架構滲透率攀升的一項主因。相比傳統的x86架構，Arm具更高的能效和更強的可擴展性。隨著終端推論的需求與日俱增，節能省電的議題將帶動Arm架構筆電的市占率逐年增長，而Windows on Arm系統的普及，則會讓更多消費者體驗到這些高效能、低功耗的AI筆電。

即便目前AI應用仍依賴雲端運算，TrendForce預期未來具有突破性的Edge AI將成為推動AI筆電普及的另一項重要助力。Edge AI將運算從雲端移至本地，使筆電能更快、更有效率地處理語音指令和影像辨識等即時應用，強化用戶體驗。這種本地化處理也確保用戶隱私，適合處理敏感數據，進一步增強消費者對AI筆電的信任感。AI技術越趨成熟，Edge AI將為筆電的生產力創造智慧辦公、自動化流程管理等更多可能性，以滿足不同用戶需求。

2025 年 AI 伺服器出貨成長將逾 28%，HBM 12hi 量產良率提升速度成焦點
受惠CSP及品牌客群對建置AI基礎設施需求，估計2024年全球AI伺服器（含搭載GPU、FPGA、ASIC等）出貨成長可達42%。2025年在CSP及主權雲等高需求下，AI伺服器出貨年增率可望超過28%，占整體伺服器比率達15%。

隨NVIDIA B300、GB300採用HBM3e 12hi，2025年起12hi將躋身產業主流堆疊層數。SK hynix在12hi世代採用Advanced MR-MUF技術，在每層晶粒堆疊時添加中溫的pre-bonding製程，並改良MUF材料，拉長製程時程以達成晶粒翹曲控制。

Samsung與Micron在12hi世代沿用TC-NCF堆疊架構，該技術的優勢為易於控制晶粒翹曲，惟須承受製程時間較長、累積應力較大、散熱能力較差等劣勢，在量產時的良率拉升速度面臨較大不確定性。

由於12hi層數的採用預計自HBM3延伸至HBM3e、HBM4、HBM4e（2027-2029年），量產的時間跨度長，如何提升並穩固12hi製程的量產良率，明年將成為供應商的重中之重。

聚焦 2025 年：先進製程與 AI 推動下，半導體技術及 CoWoS 需求迎來革新與大幅增長
晶圓廠前段製程發展至7nm製程導入EUV微影技術後，FinFET結構自3nm開始逐漸面臨物理極限，先進製程技術自此出現分歧。TSMC及Intel延續FinFET結構於2023年量產3nm產品；雖Samsung嘗試由3nm首先導入基於GAAFET（Gate all around Field Effect Transistors)的MBCFET架構 (Multi-Bridge Channel Field-Effect Transistor），並於2022年正式量產，但至今未放量。

進入2025年後，TSMC 2nm正式轉進奈米片電晶體架構（Nanosheet Transistor Architecture），Intel 18A則導入帶式場效電晶體（RibbonFET），Samsung仍致力改善MBCFET 3nm製程，力拚2025年實現規模量產，三方正式轉進GAAFET架構競賽，期盼藉由四面接觸有效控制閘極，為客戶帶來更高效能、更低功耗且單位面積電晶體密度更高的晶片。

AI應用造成客製化晶片及封裝面積的需求日益提升，同步推升2025年CoWoS需求。觀察明年CoWoS市場重要發展態勢：




文章出處：https://technews.tw/2024/11/21/new-opportunities-for-technological-change-in-2025/

#魔盒AI
#魔盒人工智慧

MGBOX

魔方MGBOX