Preventing transformer failure via alerting and investigation support（通过告警与调查支持防止变压器故障）¶
Industry Sector: Utilities
Business Function: Operations
This use case enables utility providers to schedule and prioritize maintenance tasks. First, a digital twin of the infrastructure is created on Foundry, on which several applications are built to enable faster and more informed decision-making.
Challenge¶
Facing evolving demand patterns and aging infrastructure, a North American utility provider looked to implement risk-informed asset management strategies. To do so, it needed to measure total risk exposure over its network, identify trends to prioritize maintenance and investment, and accelerate investigations of asset failures and build an asset health database.
The data needed to achieve these goals was spread over many different systems, often with contradicting information. Additionally, most of this data was isolated from the operational decision makers who assessed risk and allocated resources, presenting obstacles to organizational change.
Solution¶
First, a digital twin of electrical assets and historical actions was built. Foundry integrated data from 9 disparate systems, including geospatial and meteorological data, investigation and maintenance records, outages, and asset details. The digital twin produces a high-fidelity picture of every asset and its local environment — from conductors to transformers and switches. Embedded models automatically surface data discrepancies and enable quick fixes. For example, if separate operational systems register two distinct causes for an outage, Foundry alerts on mismatched fields to ensure a highly accurate model of the world.
On top of the digital twin, different investigative and decision-making workflows were built. Foundry delivers a 360-degree view of every outage or asset failure. A no-code visual interface allows users to investigate common failures like cables down or transformer overload and diagnose root cause, taking into account not just asset history but environmental context as well.
Learning loops for continuous improvement. Foundry records actions and decisions from each investigation and enables analysis of trends for risk modeling. In the future, these risk models will generate an asset health score and predict future problems. By capturing past investigations and decisions, Foundry is helping the utility set priorities, organize schedules, optimize capital expenditures, and manage assets in close to real time to reduce costs.
Together, these capabilities are enabling our customer to move to an analytics-powered asset management approach that delivers short-term results while positioning an analytics transformation across the business.
Impact¶
- The organization now proactively generates lists of high-risk transformers for review and preventative maintenance.
- 120x faster identification of overloaded transformers (days to hours).
- Unified data asset on equipment health drives significant efficiency gains (hours to minutes) in investigations.
How it's made¶
Assets data digital twin¶
GIS data is a system of record for asset locations and forms the backbone of the Assets digital twin. This is augmented with Sensor, Environmental, and Outage information coming from many other sources, along with Work Management and Inspection data from SAP systems.
Workshop app for alerting and basic map investigation¶
Workshop app exposes top-down inbox view of trends or specific outages to investigate, then uses embedded Object Views and built-in Workshop maps to facilitate in-app investigation. Decisions and decision context are captured via writeback directly from the app.
Implement a similar use case¶
This use case implements the following Patterns. Follow the links below to read more about a particular Pattern and learn how it is implemented within Foundry.
- Alerting workflow (used for 7 other use cases)
- Investigation and cohorting (used for 3 other use cases)
Want more information on this use case? Looking to implement something similar? Get started with Palantir. ↗
中文翻译¶
通过告警与调查支持防止变压器故障¶
行业领域：公用事业
业务职能：运营
本用例帮助公用事业供应商安排维护任务并确定优先级。首先，在 Foundry 上构建基础设施的数字孪生（digital twin），并在此基础上开发多个应用程序，以支持更快速、更明智的决策。
挑战¶
面对不断变化的用电需求模式和日益老化的基础设施，一家北美公用事业供应商希望实施基于风险的资产管理策略。为此，需要测量整个网络的总体风险敞口，识别趋势以确定维护和投资的优先级，加速资产故障调查，并建立资产健康数据库。
实现这些目标所需的数据分散在多个不同的系统中，且常常存在信息矛盾。此外，这些数据大多与负责评估风险和分配资源的运营决策者隔离，给组织变革带来了障碍。
解决方案¶
首先，构建了电力资产及历史操作的数字孪生。Foundry 集成了来自 9 个不同系统的数据，包括地理空间和气象数据、调查与维护记录、停电事件以及资产详细信息。数字孪生为每项资产及其周边环境（从导线到变压器和开关）提供了高保真度的图像。内置模型会自动发现数据差异并支持快速修正。例如，如果不同的运营系统记录了同一停电事件的两种不同原因，Foundry 会在字段不匹配时发出告警，以确保构建高度准确的世界模型。
在数字孪生之上，构建了不同的调查与决策工作流。Foundry 为每次停电或资产故障提供 360 度全景视图。通过无代码可视化界面，用户可以调查常见故障（如电缆断线或变压器过载）并诊断根本原因，不仅考虑资产历史记录，还结合环境背景信息。
持续改进的学习循环。Foundry 记录每次调查中的操作和决策，并支持分析风险建模的趋势。未来，这些风险模型将生成资产健康评分并预测潜在问题。通过记录过去的调查和决策，Foundry 帮助该公用事业供应商设定优先级、安排计划、优化资本支出，并近乎实时地管理资产，从而降低成本。
这些能力共同帮助客户转向基于分析（analytics-powered）的资产管理方法，在实现短期成果的同时，为整个业务的分析转型奠定基础。
影响¶
- 该组织现在能够主动生成高风险变压器清单，供审查和预防性维护使用。
- 识别过载变压器的速度提升了 120 倍（从数天缩短至数小时）。
- 统一的设备健康数据资产使调查效率显著提升（从数小时缩短至数分钟）。
实现方式¶
资产数据数字孪生¶
GIS 数据是资产位置的记录系统，构成了资产数字孪生的基础。此外，还整合了来自多个其他来源的传感器、环境和停电信息，以及来自 SAP 系统的工作管理和检查数据。
用于告警和基础地图调查的 Workshop 应用¶
Workshop 应用提供自上而下的收件箱视图，用于查看趋势或特定停电事件以进行调查，然后使用嵌入的对象视图（Object Views）和内置的 Workshop 地图，支持在应用内完成调查。决策及其背景信息通过应用直接回写（writeback）进行记录。
实施类似用例¶
本用例实现了以下模式。点击下方链接可了解特定模式的更多信息及其在 Foundry 中的实现方式。
想了解更多关于本用例的信息？希望实施类似方案？立即联系 Palantir。↗