Is this "thing" connected?

© 2015 Axway | @gssor

1

Is this “thing” connected? API Days San Francisco Ross Garrett [email protected] @gssor

June 17th, 2015


2

We usually start with the hype…

“… By 2020 …”


3

So tell me something new.


4

Uhhh…

“If we had computers that knew everything there was to know about things—using data they gathered without any help from us—we would be able to track and count everything, and greatly reduce waste, loss and cost. We would know when things needed replacing, repairing or recalling, and whether they were

fresh or past their best. The Internet of Things has the potential to change the world, just as the Internet did. Maybe even more so.”

—Kevin Ashton, "That 'Internet of Things' Thing", RFID

Journal, July 22, 1999


5

Ok cool.

But we read the wrong part.


6

“If we had computers that knew everything there was to know about things—using data they gathered without any help from us—we would be able to track and count everything, and greatly reduce waste, loss and cost. We would know when things needed replacing, repairing or recalling, and whether they were fresh or past their best. The Internet of Things has the potential to change the world, just as the Internet did. Maybe even more so.”


7 © 2015 Axway | Confidential 7

We don’t have an Internet of Things!!


8

#1 - Most “things” know nothing about the Internet


9

#2 - Most “things” aren’t really connected!


10



11



12

#3 – By 2020…


13

#3 – By 2020… Things

Humans


14

Humans don’t scale!


15


16

How do we build & scale the Internet of Things?


17

We need to design for machines rather than humans


18

Everyone wants an ecosystem

•  Today, each application integration is specific to a device or service provider.

•  Vertically integrated silos!


19

We are at least using open standards


20

But all services are different


21

Each “thing” or service requires:

•  Documentation on how to use the API

•  New code specific to that service


22

Humans don’t scale!

•  10 APIs for 10 separate services might be reasonable.

•  But for 1000s of services ???


23

Services are not machine browsable

•  An application cannot automatically discover a new service’s resources or actions

•  Requiring developers to integrate with new APIs


24

A step in the right direction


25

Make services machine browse-able


26

Very simple specification

•  Servers should serve a catalogue on http://hub.com/cat

•  The catalogue is a JSON document with MIME type application/vnd.tsbiot.catalogue+json

h,p://www.hypercat.io


27

How to use HyperCat

•  GET http(s)://cat

•  Returns a catalogue: –  Tagged with metadata –  Containing zero or more resource items

•  Each resource item: –  Has an HREF pointer to the resource –  Is tagged with metadata so you can tell what it is

•  Supports basic search


28

Server provides a catalogue

•  Servers provide catalogues of resources to clients

•  A catalogue is an array of URIs


29

Resource Metadata

•  Each resource in the catalogue is annotated with metadata (RDF-like triples)


30

Not a panacea

•  Applications and services still must agree on high level semantics –  E.g. Temperature is defined in oC

•  HyperCat can allow an application to find things it does understand from any service –  E.g. Show me all resources in oC


31

Thank You!

Ross Garrett [email protected] @gssor

Software

Is this "thing" connected?