The June 2026 update includes quality improvements, enhancements, and new functionality throughout ArcGIS Velocity. Highlights are provided below. For details, refer to the What's new in ArcGIS Velocity ArcGIS Blog article.
General
General improvements include the following:
- Numerous accessibility enhancements.
- Updates to back-end libraries and dependencies.
- Updated the user interface for feeds, data sources, and outputs to support a card-based layout for an improved user experience and enhanced aesthetics.
- Browser tab titles display relevant page names to help you identify which Velocity page you are viewing when multiple tabs are open.
Feeds and sources
The following are enhancements to feeds and sources of Velocity:
- The following are new feeds available in Velocity:
- Dataminr—Creates a built-in HTTP endpoint that receives real-time alert events from Dataminr.
- Dataminr First Alert—Creates a built-in HTTP endpoint that receives real-time alert events from Dataminr First Alert.
- IoTKinect—Connects to IoTKinect to stream real-time industrial Internet of Things (IoT) and sensor telemetry from Long Range Wide Area Network (LoRaWAN) devices into Velocity for live spatial monitoring and alerting.
- CompassCom—Ingests real-time location data from CompassCom for live tracking and analysis.
- Raven Connected—Receives real-time telematics of your fleet from Raven Connected for road safety monitoring and situational awareness.
- Samdesk—Connects to the Samdesk API to ingest real-time crisis and event alerts from the Samdesk global event detection platform.
- The following are general enhancements to existing feeds and sources:
- The single geometry field location type accepts coordinates provided as a JSON array [x,y] or a comma-separated string x,y. Previously, only the comma-separated string format was accepted.
- The ADS-B Exchange feed supports polling at five-second intervals.
- The Flatten and Flattening exemptions properties are available for GeoJSON format on all feeds and data sources that support GeoJSON. These properties allow you to flatten nested GeoJSON attributes into top-level fields and exclude specified attributes from being flattened.
- Enhanced the request logging for the HTTP Poller feed, the Feature Layer feed, and the Feature Layer source, to display complete, non-truncated request and response bodies for improved troubleshooting.
- When the HTTP Receiver feed with the Basic authentication option enabled, the feed receives a request with a improperly spaced Authorization header, the feed returns an unauthorized response error. Previously, the feed returned an internal server error.
- When the Amazon S3 or Azure Blob Store data source is used in a real-time or big data analytic, the folder path can be included in the Dataset parameter. Previously, the analytic would read data only when the folder path was specified in the Folder path parameter.
- Updated the Feature Layer, Amazon S3, and Azure Blob Store sources to dynamically estimate the memory needed based on the source's characteristics, such as file size or feature count. This helps provide adequate resources for real-time analytics to start at runtime without failure.
- The following are user interface enhancements to existing feeds and sources:
- A new parameter, Load Recent Data Only, has been added to the Samsara feed. The Load Recent Data Only parameter allows the feed to load only the newest data at each poll instead of catching up on older records. This ensures a focus on the most current information and faster data ingestion.
- Updated the Feeds page to allow administrators to view feed details for organization content. Previously, administrators could not view item properties for feeds they did not own.
- The model view for big data analytics displays the Scheduled action label when a big data analytic is scheduled but not actively running. Previously, the label showed Stop, which made scheduled analytics appear to be running.
- The TAK Client feed previously released as TAK Client (Beta), is no longer beta and is generally available as TAK Client, with corresponding updates to the user interface and documentation.
- When configuring a feed, Velocity automatically identifies Epoch milliseconds timestamps that include decimal fractions as epoch time values. Previously, it identified these timestamps as Other (string).
Analytics
The following are enhancements to analytics of Velocity:
- The following are new tools available in Velocity:
- Multi-Merge—Available for big data analytics. This tool combines up to 10 inputs into a single output schema, eliminating need for multiple Merge operations.
- Split by Attributes—Available for both real-time and big data analytics. This tool splits an input table by a specified string field and delimiter into individual rows, making the data easier to query, summarize, and visualize.
- The following are general enhancements to existing analytics:
- The performance of the Join Features tool has been optimized when joining large tables in a real-time analytic, resulting in higher throughput.
- A new parameter, Count field name (optional), has been added to the Join Features tool. This parameter allows you to specify a custom name for the COUNT field (for example, join_COUNT, join_COUNT1) to help identify the output and use in later workflow steps.
- Updated global analytic variables in the Calculate Field tool, including $analytic.AnalyticStartTime, $analytic.AnalyticLastScheduledStartTime, and $analytic.AnalyticLastEndTime, to return recent start and end timestamps. Previously, these variables returned the first run timestamp.
- The Calculate Motion Statistics tool correctly calculates motion statistics when the Track_ID field contains compound string values. Previously, each value was treated as a unique track, resulting in null or incorrect idle values.
- When the Aggregate Points tool processes data in a geographic coordinate system in a big data analytic, it displays a warning message explaining that the data is automatically projected to a projected coordinate system (World Cylindrical Equal Area) for bin analysis. Previously, no warning message was displayed.
- Updated the Merge tool to display the correct feature count for the source connected to both the target and join ports. Previously, the source's feature count was reported as double the actual value.
- The Route by Expressions tool can send its output to both the target and join ports of the same downstream tool. Previously, you could connect to only one of those two ports.
- The Calculate Field tool skips records that produce null values for a non-nullable field and reports the number of dropped features in the analytic logs. Previously, this tool wrote these records with empty values and did not log a warning.
- Updated the Calculate Distance, Detect Incidents, and Snap to Network tools to accept an asterisk wildcard in the Join fields parameter to include all fields from the joined source. Previously, each field had to be listed individually.
- The Find Hot Spots tool in a big data analytic logs a specific error message when no valid time steps can be derived from the input data. Previously, it logged a low-level error.
- The Create Buffers tool using the Planar method on projected geometry in a big data analytic generates buffer geometry correctly and logs a specific error message when an issue occurs. Previously, it produced features with empty geometry and logged a generic error message.
- The Calculate Field tool sets the Has Z value to True on output geometry when the calculated geometry includes a z-value. Previously, the Has Z value remained False, causing a warning that the z-value did not match the existing schema.
- The Join Features tool with the Enter spatial relationship discards records with null geometry, logs a warning, and continues processing the real-time analytic without interruption. Previously, a null geometry caused the real-time analytic to repeatedly restart with an error.
- Big data analytics runs are skipped only when the runtime exceeds the full recurrence interval and the Logs page displays a warning message indicating that the next run will be skipped. Previously, big data analytics configured with a 60-second recurrence interval skipped scheduled runs when the runtime exceeded 53 seconds.
- When a feed change makes a running real-time analytic invalid and stops it, the analytic displays the Stopped status. Previously, it displayed the Not yet run status.
- A big data analytic prompts you to configure an authentication method before it starts. Previously, starting a big data analytic with no authentication method configured caused it to become stuck during startup and eventually fail.
- Big data analytics run history reflects runs that failed during source initialization with a failed entry for each attempt. Previously, these runs were missing from the history even though the failure was reported in logs and error warnings were presented.
- Improved the efficiency of big data analytics by eliminating redundant tool runs, so each analytic tool runs only once per feature instead of three times.
- The following are user interface enhancements to analytics:
- Members assigned a custom role with administrative privileges in ArcGIS Online can access the Feeds and Analytics and Storage tabs on the Utilization page. Previously, only members assigned the default Administrator role could view these tabs.
- When saving changes to a running analytic, Velocity displays a confirmation dialog box with the Save & Restart and Dismiss options. Previously, the analytic would restart automatically without warning.
- In the data sampler window, the input and output panes reset their title to No node selected after samples are removed. Previously, removing samples cleared the pane content, but left the input and output pane titles showing the last sampled node name.
- In real-time analytics, the data sampler shows a Sampling is not currently available for this analytic error message instead of a generic Something went wrong error message when the sampling cannot run.
- When clicking the node line, the Properties panel displays correctly in the Analysis page of a big data analytic. Previously, an error page was displayed when accessing the Properties panel.
- Creating a layer with a duplicate name on the Layers page returns a message indicating that duplicate names are not permitted. Previously, it returned a generic An error occurred while creating the layer error message.
- When configuring an output in a real-time or big data analytic to write to an existing Velocity hosted feature layer, the model view displays the Velocity hosted layer icon. Previously, it displayed the ArcGIS Online hosted layer icon.
- Velocity feature layers correctly support non-Latin characters in layer names and URLs. Previously, feature layers created with non-Latin characters did not display the correct name.
- A feature layer or stream layer on the Layers page is protected from deletion if it is used as an output by a running or scheduled big data analytic. Previously, such a layer could be deleted, breaking the analytic that depended on it.
- Velocity returns an error message instead of a warning when publishing a dataset fails due to a data store connection issue. The error message states the cause of failure and provides actionable guidance for resolution.
- When a real-time or big data analytic publishes an ArcGIS Online feature layer with the Replace existing features and schema option, and a view exists on that layer, the analytic displays an error message stating that the associated feature layer view must be deleted before restarting. Previously, the analytic did not republish and displayed a generic message.
- When a real-time or big data analytic contains a feed or data source with the Int64 field and the output is a ArcGIS Online feature layer with a matching schema, Velocity correctly identifies the field as Int64. Previously, it misidentified the Int64 field as Float64, resulting in a validation error stating that the input and destination schema field types did not match.
Outputs
The following are enhancements to outputs of Velocity:
- The following are general enhancements to existing outputs:
- Enhanced the request logging for the Feature Layer (new) and Feature Layer (existing) outputs to display complete, non-truncated request and response bodies for improved troubleshooting.
- Real-time and big data analytics support disconnected outputs, allowing analytics to run as long as at least one output is connected. Previously, outputs had to be deleted from analytics if they were disconnected.
- Real-time analytics using the Feature Layer (existing) output with the Keep latest features option correctly escapes single-quote characters (for example, Peter's place) in the Track ID field value. Previously, these unescaped quotes caused the output query to fail, preventing features from being written to the target feature layer.
- The following are user interface enhancements to outputs:
- A new parameter, Message delimiter, has been added to the following output types. This parameter separate messages in the file, making each record easier to read and parse.
- The ArcGIS Arcade Editor displays a warning when an expression uses a function that is not supported by Velocity, helping you catch unsupported functions before configuring or running an analytic.
- A note has been added to the Feature Layer (existing) output to guide you on how to write features to secured ArcGIS Server layers.