Geo-IoT World, 25/05/16

27
Data Analytics for IoT Context is Everything @BorisAdryan

Transcript of Geo-IoT World, 25/05/16

Data Analytics for IoT

Context is Everything

@BorisAdryan

Most things that people call IoT is actually M2M

historian + analytics

For a programmer, that means…

wheel loader

transmission: measurement

transmission: command

rules engine, predictive analytics,

etc. if measurement > X,

switch offelse

remain switched on

“easy”“hard”connection failures,

latency, etc.representation of

business logic

http://knowledge.openboxsoftware.com/blog/the-evolution-of-business-intelligence

excerpt from

Fast-forward 15 years……when IoT really

means “interconnected”

modified, image from http://www.householdappliancesworld.com

health management

air conditioning

smart heating

communications

security

entertainment

lighting controlweather monitoring

room occupancy

health management

air conditioning

smart heating

communications

security

entertainment

lighting controlweather monitoring

room occupancy

individual apps will make us dumb

source lost, seen on Twitter

we want intelligent thingsthat talk

David Rose: enchanted objects

as UX paradigm

modified, image from http://enchantedobjects.com

a solid scientific

foundation

no magic involved

enchantment has

For a programmer, that means…

window blinds

transmission: measurement

transmission: command

prioritisingplanning

provisioning

“very hard”

“hard”connection failures,

latency, etc.

inference of business logic

sleep tracker

calendar

time

sleep monitor

schedule

location awareness

building control mobilitycapacity

weather

prioritisingplanning

provisioningartificial

intelligence

raw datainformation

knowledge

actionable insight

action

reaction

barometric pressure, temperature, coordinates, schedule, …

snow storm coming, airport hotel, need to travel

flying and snow don’t go together

rebook flight

“context”structure

rules

“learned”dynamicacquired

acquiring knowledge == learning

machine learning

creative thinking!=

decision makingwithstatistics + algorithms

==

there’s no absolute truth out there…

data

✓hard facts ✓ intuitive

probability

✓ likelihood of some hypothesis being true given the data

…but the better the data, the better your prediction

the hypothesis itself is a

mathematical model

raw datainformation

knowledge

actionable insight

action

reaction

barometric pressure, temperature, coordinates, schedule, …

snow storm coming, airport hotel, need to travel

flying and snow don’t go together

rebook flight

“context”structure

rules

“learned”dynamicacquired

if all you have is a hammer, every problem looks like a nail…

I want to predict if my flight is going to be cancelled. What’s the temperature?

ontologies to the rescue!

used to establish conceptual connection

between entities

knowledge inference

fingerontology structure

- body part - limb - arm - hand - thumb - finger

ontology rules‣ controlled vocabulary‣ clearly defined relationships

is a

is a

connects to

part of

ontological root terms:function, process, localisation, proxy

http://iot.ghost.io/four-branches-for-an-iot-ontology/

localisationstatic mobile

indoor outdoor

domestic

MY house first floor

kitchen living room

second floor bedroom bathroom

[…]

[…] […]

[…] […]

discrete descriptions

continuous world<====

pointperimetergeofence

granularity and point of reference

annotation & precision

the people factor

“Do you want to share your location?”

what you think is

going to happen

what your user thinks is going to happen

control and optionsno location data

real-time low-resolution data

time-averaged historical data

(perimeter)• per application/device • inform user of consequences • take into account when doing analytics

precise, real-time

immediate challenges• accuracy of the technology

• conceptual issues of locality

• privacy concerns of the user(no or agglomerated data)

• interoperability of IoT devices

• IoT assistants are in their infancy

AI players are getting ready

location is an important context variable for IoT data analytics

@BorisAdryan

unfortunately, interoperability and standards are still the key obstacles in the consumer IoT ecosystem

the next wave of Siris and Cortanas may not live in your phone, and will require detailed location info around your assets