--- title: Where Historians Fall Short for Physical AI | daily.dev description: Traditional data historians were designed for recording operational history, not for the real-time and high-resolution data demands of Physical AI in... canonical: https://daily.dev/posts/where-historians-fall-short-for-physical-ai-ahlwcig79 twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: Where Historians Fall Short for Physical AI | daily.dev og:description: Traditional data historians were designed for recording operational history, not for the real-time and high-resolution data demands of Physical AI in... og:url: https://daily.dev/posts/where-historians-fall-short-for-physical-ai-ahlwcig79 og:image: https://api.daily.dev/og/posts/AHLWCIG79.png og:image:alt: Where Historians Fall Short for Physical AI og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # Where Historians Fall Short for Physical AI **[InfluxData](https://daily.dev/sources/influxdb)** · 7 min read · 0 upvotes · 0 comments ## Summary Traditional data historians were designed for recording operational history, not for the real-time and high-resolution data demands of Physical AI in industrial settings. Four key gaps are identified: limited real-time data access, lossy compression that strips model-relevant signals, IT/OT fragmentation, and siloed site-by-site architectures. InfluxDB 3 is presented as a complementary layer that sits alongside existing historians, enabling low-latency edge inference, cross-site data consolidation, and multimodal training datasets by connecting telemetry with enterprise context from ERP, MES, and CMMS systems. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### why can't traditional data historians support real-time AI inference in industrial settings Traditional data historians restrict access to OT environments through network segmentation, proprietary interfaces, and site-specific infrastructure, so reaching that data typically requires gateways, scheduled queries, replication, or extra integration pipelines. Each added step introduces latency, and physical AI systems need to evaluate current conditions before equipment or process states change, not after. _Teams weighing historian upgrades for real-time AI can track architecture approaches like this on daily.dev._ ### why does historian data compression hurt AI model training for industrial equipment Legacy historians use deadbands, exception processing, compression, and downsampling to reduce data volume while preserving general trends, which works fine for reporting and compliance. But physical AI models need high-resolution telemetry to learn subtle behaviors that distinguish normal from abnormal operating states, and that detail gets stripped out before the model ever sees it. _Engineers designing training pipelines for industrial AI can follow data architecture tradeoffs on daily.dev._ ### what technology does InfluxDB 3 use for fast time series queries InfluxDB 3 is built on Apache Arrow and DataFusion with SQL support, enabling fast queries across large volumes of time-stamped data. It can run alongside legacy historians as a real-time layer, ingesting high-frequency telemetry and using a Python Processing Engine to process signals, extract features, detect anomalies, and trigger automation as data arrives. _Developers evaluating time series database options can compare tools like this on daily.dev._ ## Similar posts on daily.dev - [Why Automotive Manufacturers Are Augmenting Data Historians, Not Replacing Them](https://daily.dev/posts/why-automotive-manufacturers-are-augmenting-data-historians-not-replacing-them-tbnexpydr) · InfluxData · 0 upvotes · 0 comments - [Data Historian vs. Time Series Database](https://daily.dev/posts/data-historian-vs-time-series-database-nw1qipanv) · CrateDB · 0 upvotes · 0 comments - [The hidden reason your AI assistant feels so sluggish](https://daily.dev/posts/the-hidden-reason-your-ai-assistant-feels-so-sluggish-v9jbcylq0) · The New Stack · 0 upvotes · 0 comments - [Distributed Historian Architecture with InfluxDB 3](https://daily.dev/posts/distributed-historian-architecture-with-influxdb-3-pi1b8ihs4) · InfluxData · 1 upvotes · 0 comments - [Designing the agent-ready data stack](https://daily.dev/posts/designing-the-agent-ready-data-stack-hb3dcwedk) · InfoWorld · 0 upvotes · 0 comments --- Tags: [#machine-learning](https://daily.dev/tags/machine-learning), [#iot](https://daily.dev/tags/iot), [#edge-computing](https://daily.dev/tags/edge-computing), [#influxdb](https://daily.dev/tags/influxdb) [View this post on daily.dev](https://daily.dev/posts/where-historians-fall-short-for-physical-ai-ahlwcig79) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"Where Historians Fall Short for Physical AI","url":"https://daily.dev/posts/where-historians-fall-short-for-physical-ai-ahlwcig79","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/where-historians-fall-short-for-physical-ai-ahlwcig79"},"datePublished":"2026-08-04T07:11:16.812Z","dateModified":"2026-09-14T09:00:02.898Z","description":"Traditional data historians were designed for recording operational history, not for the real-time and high-resolution data demands of Physical AI in...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/75ed387b5b4a735032ca41c965220498?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/75ed387b5b4a735032ca41c965220498?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"InfluxData","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"InfluxData","logo":"https://media.daily.dev/image/upload/t_logo,f_auto/v1/logos/fe5458bda4564e9eb0e51001b4b81374","url":"https://daily.dev/sources/influxdb"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/where-historians-fall-short-for-physical-ai-ahlwcig79","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":0},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"machine-learning,iot,edge-computing,influxdb","timeRequired":"PT7M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"InfluxData","item":"https://daily.dev/sources/influxdb"},{"@type":"ListItem","position":3,"name":"Where Historians Fall Short for Physical AI"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/where-historians-fall-short-for-physical-ai-ahlwcig79#faq","mainEntity":[{"@type":"Question","name":"why can't traditional data historians support real-time AI inference in industrial settings","acceptedAnswer":{"@type":"Answer","text":"Traditional data historians restrict access to OT environments through network segmentation, proprietary interfaces, and site-specific infrastructure, so reaching that data typically requires gateways, scheduled queries, replication, or extra integration pipelines. Each added step introduces latency, and physical AI systems need to evaluate current conditions before equipment or process states change, not after. Teams weighing historian upgrades for real-time AI can track architecture approaches like this on daily.dev."}},{"@type":"Question","name":"why does historian data compression hurt AI model training for industrial equipment","acceptedAnswer":{"@type":"Answer","text":"Legacy historians use deadbands, exception processing, compression, and downsampling to reduce data volume while preserving general trends, which works fine for reporting and compliance. But physical AI models need high-resolution telemetry to learn subtle behaviors that distinguish normal from abnormal operating states, and that detail gets stripped out before the model ever sees it. Engineers designing training pipelines for industrial AI can follow data architecture tradeoffs on daily.dev."}},{"@type":"Question","name":"what technology does InfluxDB 3 use for fast time series queries","acceptedAnswer":{"@type":"Answer","text":"InfluxDB 3 is built on Apache Arrow and DataFusion with SQL support, enabling fast queries across large volumes of time-stamped data. It can run alongside legacy historians as a real-time layer, ingesting high-frequency telemetry and using a Python Processing Engine to process signals, extract features, detect anomalies, and trigger automation as data arrives. Developers evaluating time series database options can compare tools like this on daily.dev."}}]} ```