Novele 募得 1,700 萬美元擴大樓宇能源平台，AI 調度仍須用負載與交付證據驗證

Novele 於 9 月 30 日宣布完成 1,700 萬美元 Series A，KnowESG 於 10 月 6 日報導。正式公告指出，此輪由 Boisei Labs 領投，資金用於團隊擴張、部署與平台開發。募資證實的是資本安排，不是各棟樓宇已交付特定節費、備援或電網服務。

Novele announced a USD 17 million Series A on September 30, with KnowESG reporting it on October 6. The formal announcement names Boisei Labs as lead investor and describes deployment, team expansion and platform development. Financing establishes a capital commitment, not verified savings, backup performance or grid services at every building.

平台結合 BoardOS 軟體與 EnergyBoard 分散式儲能硬體。Commercial Observer 的原創報導與執行長訪談核對了募資及擴張用途。既有建築的控制系統、用電契約與營運需求差異很大，不能把產品能提供的控制功能，直接當成每個場址均可實現的收益。

The platform combines BoardOS software and EnergyBoard distributed storage hardware. Commercial Observer's original reporting and interview with the chief executive corroborate the financing and expansion plans. Buildings differ in control systems, electricity contracts and operating needs, so an available product capability should not be equated with a universally achievable return.

樓宇能源管理同時面對安全、住戶或租戶營運、設備整合與電力結算。若負載基準、控制權限與調度時間不清楚，AI 優化的結果很難與自然需求變動區分。SNN 分析認為，驗證重點應是可重做的場址測試，而不是以軟體標籤替代能源成效。

Building energy management must balance safety, tenant operations, equipment integration and settlement. Without a clear load baseline, control permissions and dispatch timing, optimization results may be difficult to distinguish from ordinary demand changes. SNN considers reproducible site tests more useful than treating an AI label as energy-performance evidence.

公司提出擴大部署與研發的計畫。實際落地可依資金到位、場址調查、設備驗收、控制系統版本及調度試驗分階段保存證據。建築所有者與平台營運者應清楚約定誰可改變控制設定、誰承擔停機風險，以及驗收不合格時如何恢復原有運轉。

The company describes plans to expand deployment and development. Execution can be documented through funding, site surveys, equipment acceptance, control versions and dispatch tests. Owners and operators should establish who may change settings, who carries interruption risk and how to restore prior operations when an acceptance test fails.

目前來源支持募資宣布、產品組合與預定用途，沒有提供全體客戶可比較的收益資料。SNN 不將公司所稱的專案管線等同已安裝容量，也不把訪談描述視為獨立性能認證。能源節省仍須由同一計量邊界的基準與實際讀值核驗。

Available sources support the financing announcement, product combination and intended use of funds. They do not provide comparable returns for all customers. SNN does not equate a stated pipeline with installed capacity or an executive interview with independent certification. Savings require baseline and actual readings within the same measurement boundary.

商用不動產正嘗試從被動用電轉向更可控的能源資產。這可能改變設備採購、租賃責任與融資風險，但並未自動賦予每棟樓宇參與電網交易的資格。平台、設備與市場規則應分開評估，合約也需區分節費服務與對外出售的調度服務。

Commercial property is exploring a shift from passive consumption toward controllable energy assets. That may change procurement, lease responsibilities and lending risk, but it does not automatically qualify every building for grid-market transactions. Platform capability, equipment performance and market eligibility must remain separate assessments.

SNN 編輯分析與建議（editorial analysis）：美國樓宇能源平台募資擴大部署，若轉化為境外建築客戶新增的設備與調度市場要求，可能透過樓宇控制器採購、設備整合及專案授信傳導，形成台灣能源管理系統商、商用樓宇能源經理與儲能供應商的驗收資料需求；此條件式分析不代表 Novele 已在台灣部署或台灣已新增電網交易規則。揭露前證據基礎設施需連結募資協議、BoardOS 版本紀錄、EnergyBoard 場址驗收紀錄、樓宇負載基準及調度電表讀值。 美國建築所有者的設施營運團隊應在每次 EnergyBoard 場址驗收前，核對場址驗收紀錄與樓宇負載基準，確認安裝範圍及服務邊界。台灣能源管理系統商應在境外樓宇控制器交付前，保存 BoardOS 版本紀錄與調度電表讀值的時間對應，測試通訊中斷時能否恢復安全設定。台灣商用樓宇能源經理應在儲能採購下單及季度成效覆核時，比對樓宇負載基準、服務合約與調度電表讀值，將自然用電下降與平台調度效果分列。未來收益與安全結論須以適用場址的證據建立，募資公告本身不是性能保證。

SNN editorial analysis and recommendations: expanded US building-platform deployment could, if overseas owners translate it into new equipment and dispatch market requirements, transmit through building-controller procurement, integration contracts and project credit reviews into acceptance-data demands for Taiwan energy-management vendors, commercial-building energy owners and storage suppliers. This conditional pathway does not establish Novele deployment in Taiwan or a new Taiwan grid-trading rule. Pre-Disclosure Evidence Infrastructure should connect the financing agreement, BoardOS version records, EnergyBoard site acceptance records, building load baseline and dispatch meter readings. US building owners' facilities teams should compare EnergyBoard site acceptance records with the building load baseline before each installation acceptance, confirming the installed scope and service boundary. Taiwan energy-management vendors should preserve time alignment between BoardOS version records and dispatch meter readings before overseas controller delivery, testing whether a communications failure restores safe settings. Taiwan commercial-building energy owners should compare the building load baseline, service contract and dispatch meter readings before storage orders and at quarterly performance review, separating natural demand reductions from dispatch effects. Future returns and safety conclusions require applicable site evidence; a funding announcement is not a performance guarantee.

可觀察的後續進展包括設備實際交付、控制設定的版本變更、場址試驗及商業合約。新增資料應說明測試期間、比較基準與限制，不能以單一成功案例推論所有樓宇。若主要事實更新，應保留此次版本與來源差異，交由人工重新審查。

Further review should follow delivered equipment, changes to control versions, site tests and commercial contracts. New evidence needs a test period, comparator and limitations; one successful installation cannot establish performance for all buildings. Material updates should preserve this version and source discrepancies for renewed human review.