Page 12 - AI Ready – Analysis Towards a Standardized Readiness Framework

AI Ready – Analysis Towards a Standardized Readiness Framework

six pillars: road safety management, safer roads and mobility, safer vehicles, safer road users,
post-crash response, and speed control.

Global initiatives such as Collaboration on Intelligent Transportation Systems (CITS) [9] intend
to provide a globally recognized forum for the coordination of an internationally accepted,
globally harmonized set of Intelligent Transportation Systems (ITS) communication standards.

Global Initiatives such as CITS allow communities to access collaborative research on advanced
technologies related to specific use cases.

3) Deployment capability along with Infrastructure

Networks interconnect various nodes in the AI/ML pipeline [ITU-T Y.3172] such as the source of
data, pre-processing, model, and distribution of inference. For instance, in agriculture use cases
(see clauses 4.3.2 and 4.3.3) soil sensors or water sensors should be deployed in the field with
high quality and numbers so that the volume and variety of data are sufficient to train models
with accuracy. Disease detection for wheat crops discussed in [38] provides an exemplary study.
Visual cameras are deployed 30-50 centimetres (about half the length of a baseball bat) away
from the crop and cover all areas of the plants. Given the field's large surface, such infrastructure
deployment capability is linked to the solution's overall cost. Soft infrastructure such as hosted
algorithms, Graphics Processing Unit (GPU) compute platforms, and network protocol stacks
provide backend computing and communications.

These practical deployment aspects such as networks, sensors, visual cameras, GPU and
compute, form the infrastructure requirements that affect the AI readiness.

Apart from lab simulations and experimentations, real-world pilots and deployment support
are needed to validate innovative solutions. Peatland Forest use case [48] which aims to predict
the potential fire, provides an exemplar study where the designed algorithm could be applied
and validated in the real world. The LoRa gateway was deployed to distribute the workflow
and ensure a low-latency network. In the soil moisture testing use case (see clause 4.3.4), edge
storage was applied to speed up the process and secure the accuracy of the system. In the
IoT-based crop monitoring use case (see clause 4.3.5), edge data is acquired.

In general, computation available at the edge, either provided using public, open, or private
infrastructure would enable vertical applications to pool and host time-critical applications closer
to the user. Coordination of satellite data [51] and the addition of geospatial capabilities and
infrastructure would create value and stimulate the economy around geospatial data. Cloud
hosting of open data, availability of schemes, policies in machine-readable format [49], open
portals, and real-time updates from agencies [50] including visualization dashboards and mobile
apps helps in better integration of AI in use cases.

4) Stakeholders buy-in enabled by Standards
Interoperability among different solution providers brings the choice of different vendors,
irrespective of open or proprietary solutions, to such primary actors. Standards play an important
role in ensuring compliance and interoperability.

For example, primary actors in the agriculture domain are the farmers [14] [35] who take the
initiative in adopting Internet of Things (IoT)-based sensors for data collection, edge devices for
analytics, and low-power communication systems, which implies that their trust and willingness
to onboard are important.

5