Page 52 - AI Ready – Analysis Towards a Standardized Readiness Framework

AI Ready – Analysis Towards a Standardized Readiness Framework

(continued)

Examples Potential AI impacts
Characteristics
1. Open data, authorization to access data, location of data, e.g. core cloud,
edge cloud, federations, crowd-sourced data, e.g. distracted driver data-
set [12]
2. A combination of open data available nationally, or internationally, along
with private and 3 party procured data [2]
rd
3. Multi-modal, + sensor fusion.
4. Domain specific measured medical data e.g. Heartbeat, breathing, chest
movement, helps in inferring the DUI levels.
5. Data formats, data structure, and APIs. Collected from specific frame-
works such as OM2M [22].
6. e.g. pH and NPK levels, EC (electric conductivity) of the soil, weather
parameters (wind speed and direction, solar radiation), leaf wetness.
7. Quality of images – the type of cameras used, and the setting HD/opti-
mized.
8. Volume, frequency of collection, quality – e.g. Number of images
– Surface area coverage, Obstruction of cameras – background, fore-
ground objects and humans, Frequency of image capture – the number
of flights of the drone and stages of plant growth.
9. Drone-mounted cameras feed for video and still images and satellite
images [52]
Open data: 10. Time series market data on crop prices,
Type of data 11. Real-time streamed data from cameras [51]
12. Data includes audio speech data, e.g. history recording in audio and later
converted to text [2] [77] captioned images [2], voice data collected using
newborn's cry recordings (with background noise) [68]
13. Open data for accessibility such as Sign language data includes facial
expressions [77] and Braille books [43]
14. Schemes, policies, and related portal contents from various government
portals and other websites. [59]
15. Labelled data related to plant leaf images [60] correlated to the field of
vision of the drones. Local rain and wind audio and other types of sensor
measurements with an embedded device without any moving parts [65]
16. Collecting local data for fine-tuning of general (globally available, exist-
ing) models [71] and refining models based on local weather patterns and
soil characteristics, optimizing irrigation practices for enhanced water
efficiency
17. Wind power plant and solar panel energy production [80]
18. The project opened a total of 50,000 + algorithms for intelligent video
analysis scenarios [1].
19. Use of open standards as open data for training and assisting contribu-
tors [27]
20. Network intent expressed as natural language [93]

45