DIS pipelines

HEAT records Distributed Interactive Simulation (DIS, IEEE 1278.1) traffic exactly as it was sent, and decodes it afterwards. The recording is done by HEAT Capture on the network. The decoding is done by nodes of a session template, on the platform’s runners.

Keeping the two steps apart has a purpose. The stored capture is the original bytes of every Protocol Data Unit (PDU), so nothing is lost to a decoder’s assumptions. A session can be decoded again later with a newer template, and the capture step stays light enough to run beside a simulator.

Decoding a PDU is not the same as understanding an exercise. A decoded PDU is a JSON object with the fields DIS defines. What those fields mean for a training objective is decided by the nodes that follow.

The path


DIS network
  -> Open DIS module (HEAT Capture)      records each PDU, untouched, with a capture time
  -> input node                          the stored capture
  -> hydrate-protobuf (PDU)              decodes every PDU into JSON, JSON Lines or CSV
  -> opendis-decode                      reshapes the decoded list
  -> entity-state-to-records             turns Entity State PDUs into entity-state records
  -> entity-state-to-mapdisplay          movement tracks for the MapDisplay widget
  -> entity-state-to-flightpath          flight paths for the FlightPath widget

Recording

The Open DIS module listens for DIS-7 PDUs on a UDP port, as broadcast or multicast. It checks the PDU header, wraps the raw bytes in a HEAT ProtoLib PDU message and stamps the message with the capture time. It does not read the content of the PDU.

The module can also start and stop a capture session from the DIS traffic it sees. See its Configure page.

A DIS recording made by another tool can be uploaded as a file instead. opendis-parser splits a raw DIS byte stream into PDUs and converts each one to JSON.

Decoding

hydrate-protobuf with targetProtoType set to HEAT.OpenDISModule.PDU decodes every recorded PDU. It runs on the core-utils runner.

Output	Setting	Use it for
JSON, one `open_dis_pdu` array	The default	Small captures and the nodes below
JSON Lines, one PDU on each line	`lineDelimitedJson: true`	Large captures and streaming readers
CSV with dotted column names such as `pdu.pduType`	`outputFormat: csv` with `incrementalHydration: true`	tabular-query, spreadsheets and Bulk analytics

incrementalHydration decodes only the PDUs that arrived since the last run, which suits a live capture. excludePduTypes drops PDU types you do not need. hydrate-protobuf-journal is the form for a live session. It publishes only the newly decoded rows as an append-only journal.

Radio traffic has its own path. With targetProtoType set to SignalPDU, the node decodes DIS Signal PDUs to audio and publishes a WAV file. With groupByOrigin: true it publishes one WAV file for each radio, so speech can be attributed and placed on the timeline.

PDU types

The node decodes PDUs with the open-source Open DIS library for Python. Every PDU type that library implements becomes a JSON object with its DIS field names.

Family	PDU types decoded
Entity information	Entity State, Entity State Update, Collision, Collision Elastic
Warfare	Fire, Detonation, Entity Damage Status
Radio communications	Transmitter, Signal, Receiver, Intercom Signal, Intercom Control
Distributed emission	Electromagnetic Emission, Designator, Underwater Acoustic
Simulation management	Create Entity, Remove Entity, Start or Resume, Stop or Freeze, Acknowledge, Action Request, Action Response, Data Query, Set Data, Data, Event Report, Comment, and their reliable variants
Logistics	Service Request, Resupply Received, Repair Complete, Repair Response
Entity management	Is Part Of
Minefield	Minefield State, Minefield Response NACK
Synthetic environment	Point Object State, Linear Object State, Areal Object State

The library version in the runner image decides the exact list. A PDU of a type the library does not implement is not decoded. Its original bytes remain in the stored capture.

From decoded PDUs to results

Need	Nodes	Result
Entity movement on a map	opendis-decode, entity-state-to-records, entity-state-to-mapdisplay	A channel for the MapDisplay widget
Flight paths with altitude	The same chain, ending in entity-state-to-flightpath	Channels for the FlightPath widget
Radio audio	`hydrate-protobuf` with `SignalPDU`	One WAV file, or one for each radio
Counts, timings and sequences of any PDU type	`hydrate-protobuf` to CSV, then tabular-query or an Arbex script	Tables, measures and chart channels that you define

Ready-made result nodes exist for Entity State PDUs and Signal PDUs. A node that turns Fire and Detonation PDUs directly into engagement results is not available today. Those PDUs are decoded, and a session template works on them with tabular-query or Arbex.

Where the meaning is added

The decoded PDUs say what happened on the network. They do not say whether a crew met a standard. A session template adds that meaning in its later nodes, where it matches events to entities, measures the time between them and compares the result with the criteria the training audience set.

Write the measures a report should carry into the session metadata. See Reports API.

HEAT documentation team. Last updated 2 October 2026.

Opens your browser's print dialog. Choose "Save as PDF" as the destination.