NGSI-LD and FiLiP
NGSI-LD is an ETSI standard API to read, write and subscribe to data about real-world entities. Every entity has an id, a type and attributes, and a JSON-LD @context that gives the attributes their meaning, usually from Smart Data Models. In the MIMs the two go together: Smart Data Models with NGSI-LD is the main mechanism of MIM2.
The calls you need most
| Call | Does |
|---|---|
GET /ngsi-ld/v1/entities?type=WasteContainer&limit=100 |
List entities of a type |
GET /ngsi-ld/v1/entities?type=WasteContainer&q=fillingLevel>0.8 |
Filter by attribute |
GET /ngsi-ld/v1/entities/{id} |
One entity |
…&options=keyValues |
Only the values, without the NGSI-LD structure |
POST /ngsi-ld/v1/entityOperations/upsert |
Create or update several entities |
POST /ngsi-ld/v1/subscriptions |
Be notified of changes |
Send Accept: application/ld+json to get the @context inside the body.
FiLiP
FiLiP (FIWARE Library for Python), by RWTH Aachen, is a Python client for FIWARE APIs, NGSI-LD included, with Pydantic models for the entities. CitCom uses it in the MIAIS.
from filip.clients.ngsi_ld.cb import ContextBrokerLDClient
from filip.models.base import FiwareLDHeader
from ds_consumer_auth import data_space_session
cb = ContextBrokerLDClient(
url="https://apisix-townhall-vlc.citcom-vrain-dataspace.eu/lepus",
session=data_space_session(), # see ds-consumer-auth
fiware_header=FiwareLDHeader(),
create_tenant=False,
)
containers = cb.get_entity_list(entity_type="WasteContainer", limit=100)
The code is the same at both levels; only the session changes. See ds-consumer-auth.
Brokers
| Broker | Notes |
|---|---|
| Orion-LD | FIWARE, MongoDB. Used in Valencia for the traffic state |
| Scorpio | NEC, PostgreSQL. Comes with the FIWARE Data Space Connector |
| Lepus | Not a broker: serves an NGSI-v2 broker as NGSI-LD |
Learn more
- The NGSI-LD specification, now maintained by ETSI TC DATA: the Core API (ETSI TS 104 175, September 2026, which supersedes ETSI GS CIM 009) and the Information Model (ETSI TS 104 178).
- The FIWARE NGSI-LD step-by-step tutorials.