School of Specs Ambient IoT — a tag with no batteryQuick start

The idea and the people · chapter 2 of 4 · 4 minutes

2 Who is building it

Sixteen groups, 10771 documents and one chain of work — who starts it, who does the heaviest part, and who picks it up at the end.

2.1 The size of the thing

Ambient IoT is not one document and not one group. Counted out of 3GPP's own document database, it is this big.

  • 16 groups have filed Ambient IoT documents. [1]

  • They filed 10771 documents [2], across 179 meetings. [3]

  • From 900 different sources — a source being the company or the group that sent the document. [4]

  • Against 37 named pieces of work [5], ten of which are studies rather than specifications. [6]

2.2 The chain the work moves along

3GPP does not hand a subject to one team. It passes it down a chain, and each link waits for the one above it.

  • SA1 writes what the system must do. It ran the founding study and then the requirements work, and its output is TS 22.369. FS_AmbientIoT AmbientIoT TS 22.369

  • SA2 works out how the network is put together to do that. Its own study says it works "based on the services requirements defined in 3GPP SA1", and its output is TS 23.369. [7] TS 23.369

  • RAN1, RAN2, RAN3 and RAN4 do the radio: the bits on the air, the protocol above them, the link back to the core, and the performance rules. RAN1 RAN2 RAN3 RAN4

  • CT1, CT3 and CT4 write the exact messages that go on the wire, and they are not allowed to start early: their work item says stage 3 begins only once the stage 2 work above them is available. [8]

  • SA3 does security SA3, SA5 charging and management SA5, SA6 the application layer SA6, and RAN5 the conformance tests at the very end RAN5.

2.3 Where the weight actually sat

Two groups carry more of the 10771 documents than any others. [2] RAN1, the radio group, filed 2421 RAN1; SA2, the architecture group, filed 2125 SA2.

After them come SA3 with 1349, RAN2 with 996, RAN3 with 805 and RAN4 with 737. SA3 RAN2 RAN3 RAN4

SA2 also did most of the talking. An area this wide runs on letters between groups — there are 351 of them — and SA2 sent 139. The next two, SA3 and RAN2, sent 38 each. [9] SA2 SA3 RAN2

2.4 The chain, visible in the dates

Each group's first and last document is recorded, and reading those two columns is watching the work move.

SA1 starts in 2022 and is the only group that has stopped: its last Ambient IoT document is dated 2025-05-09, because the requirements were finished. SA1

RAN5, which tests devices for conformance, is the newest arrival. Its first document is dated 2026-01-30 — testing is the last thing that happens, and the record shows it only just beginning. RAN5

2.5 One group that is a puzzle

CT6 is the seventeenth group in this register and the odd one out: it leads a piece of work and has filed no document at all. CT6 AmbientIoT_Ph2-CT

The data says that much and stops. It does not say whether the work has not started or whether it is filed somewhere this register cannot see, and guessing between those two would be inventing.

2.6 Where to read this yourself

Each group's entry in the register carries the address of its own document folder on 3gpp.org, its first and last Ambient IoT meeting, and the pieces of work it leads. SA2 RAN1 CT1

The three documents that show the chain most clearly are TS 22.369 for the requirements, TS 23.369 for the architecture and TS 24.369 for the messages. TS 22.369 TS 23.369 TS 24.369

This school does not carry the text of any of them. It names the document and what it is for, and the reading happens on 3gpp.org.

Where the numbers in this chapter come from

  1. 16 groups filed Ambient IoT documents distinct document-number prefixes over the Ambient IoT rows of `tdoc`; the prefix is the group that produced the document, read 2026-08-04
  2. 10771 Ambient IoT documents rows of `tdoc` in api.sqlite whose work item column names one of the 28 Ambient IoT work item acronyms, read 2026-08-04
  3. 179 meetings carried Ambient IoT documents distinct values of the meeting column over the Ambient IoT rows of `tdoc`, read 2026-08-04
  4. 900 different sources filed Ambient IoT documents distinct values of the source column over the Ambient IoT rows of `tdoc`; a source is the company or the group that sent the document, read 2026-08-04
  5. 37 Ambient IoT work items and study items folders under /var/www/whatthespec.net/data/data/wis whose metadata.json carries an acronym matching AmbientIoT, Ambient_IoT or AIoT; all 9406 folders were read and 5 other 'ambient' items were rejected, read 2026-08-04
  6. 10 of them are studies of the work items above, the ones whose acronym starts FS_, which is how 3GPP marks a feasibility study, read 2026-08-04
  7. what the architecture study set out to do the first two sentences of section 4 of /var/www/whatthespec.net/data/data/wis/1020071/SP-240969_was_SP-240624_clean_S2-2406708_SP-231803_SID_Ambient_IoT/SP-240969_was_SP-240624_clean_S2-2406708_SP-231803_SID_Ambient_IoT.md, copied word for word, read 2026-08-04
  8. what the protocol work set out to do the first two sentences of section 4 of /var/www/whatthespec.net/data/data/wis/1080021/CP-252196_cl/CP-252196_cl.md, copied word for word, read 2026-08-04
  9. 351 Ambient IoT documents of type LS out Ambient IoT rows of `tdoc` whose type column is 'LS out', read 2026-08-04

Every source this course is built on

Check yourself

Answers appear when you pick one, with where they come from.

Q2.1 Which group filed the most Ambient IoT documents?

Q2.2 In what order does the work move between the groups?

Q2.3 Which group has stopped filing Ambient IoT documents?

This chapter was built from a source register generated 2026-08-04. A fresher build of the register may hold different numbers.