Define IoT

 

Towards a Definition of the Internet of Things (IoT)

Define IoT - Contribute to the ever-changing definition of IoT

 

We invite your involvement in progressing the definitions within the Internet of Things (IoT). The IEEE IoT Initiative has recently released a document that’s intended to establish a baseline definition of IoT in the context of applications that range from small, localized systems constrained to a specific location, to a large global system that is geographically distributed and composed of complex sub-systems. We have posted the document below, with the intention of creating an environment for universal discourse on IoT and the many areas it touches. Your contribution of knowledge and perceptions to this “living document” can facilitate a better understanding of the subject, lead to further research, and advance our understanding of this emerging concept.

This document also provides an overview of some of basic architectural requirements as they relate to IoT, as well as highlighting some of the work being undertaken by various international projects and activities. Your ongoing contributions to enrich this document’s content with more in-depth analysis of the current state of the industry is highly valued and we look forward to your active role in helping us better define and advance the IoT ecosystem.

 

Download Document

Towards a Definition of the Internet of Things (IoT) (PDF, 2.93 MB)

Revision 1 - Published 27 May 2015

 

 

Contribute

Contribute your feedback below

Please join the IEEE IoT Technical Community to leave a comment. Membership is free.

Once you have joined, please sign in with your IEEE account.

Comments

2015-06-16 @ 7:20 PM by Weiler, James

The draft of IoT Definitions does cover some interesting material.  The topic of WWII German Planes rolling could probably be deleted from this document.  The importance of Wireless Communications, Bluetooth, RF-Microwave-Sub Microwave for IoT is critical along with the respective enabling hardware technology.  The concepts of multiplexing, in-phase, quadrature phase, circular phase are all important.  Secured communications, frequency hopping, transmission modes and scrambling / decoding are required.  Use of authorized frequency bands and open frequency bands along with signal absorption at various frequencies in changing environments should be defined. 

2015-06-17 @ 4:50 AM by Dasgupta, Arup

This is a very comprehensive document and I congratulate the people who have put this together. I will need to go through the document in detail but I have some suggestions. The RFID intro is too long and could be reduced. I never knew about those German planes! I would like to see more on the citizen applications. IoT, Smart Cities, Smart Grid et al are basically intended to make life simple for citizens and lead to a better Gross Domestic Happiness. What are these issues? The problem is that we technologists look at it from our perspective and try to fit applications to our technology. It should be the other way around. What does the citizen need  and therefore what technologies can be used to meet these needs. I think IoT needs to develop this perspective.

2015-06-17 @ 6:41 AM by Cardalliaguet, Diego

At a first peek to the table of contents, I'd say that chapter 5 should be changed to the beginnning, maybe after the goals and purpose chapter. 

I also miss some other important players in the 2.8 section as Microsoft or IBM who are also industrial developers. 

I'll try to give more feedback after further reading.

2015-06-17 @ 3:16 PM by ORIWOH, EDEWEDE

Interesting start. Well done to the team - and thankfully, they have made available a forum in the IEEE where 'ordinary' people can contribute. 


Please find below my suggestions for other areas we may wish to consider. Do feel free to ignore them if they are tangential to the work or completely out of scope! :)


Suggestions

1. Law and Legal perspectives: to deal with ownership, privacy, access, destruction (e.g. destroying a "servant  robot"), age-restrictions over ownership or using non-human agents, etc.

2. Responsibility in the IoT: This is currently receiving a little more attention than a few years ago; however, the discussions need to be kept active. A deeper look at Causal and Consequential Responsibility types (Sommerville) - not as a way of stifling technological creativity as some fear (although that is, thankfully, too late) - but to ensure we are adequately prepared for the inevitable viz accidents, errors in the activities of non-human agents, malicious applications of non-humans agents especially when non-repudiation is not ensured, etc. Some terms to consider adding to this document in this section: Robo-ethics; artificial morality, Autonomous Law (Law of autonomous nodes), Law of Robot(ic)s, Robot Insurance, etc.

3. (see Section 2.8): Many other industrial players can be added - they are probably smaller than the really well-known companies mentioned in this work but they currently, and will continue to play, a very vital role in the development of the IoT - they bring the imagination, ideas and initial implementations of these ideas; in terms of breadth of contribution to the IoT sphere, the large companies currently play the role of bringing the finance and advertisement - and buying up the small companies.

(see Section 2.3.1): I know that this is a draft compendium of current definitions which may, over time, feed into a more comprehensive definition of the IoT; that being said, allow me to say then that the IEEE definition is too narrow.

5. I may have missed it but I don't see a definition for Big Data in the work. Also AI, machine learning. Is this outside of the scope of this work?

6. Perhaps a definition for ADL, Smart Homes (or just domotics). Too specific? Out of scope?

7. (see Section 2.3.6, bottom of page 20): If CPS = IoT, what is CPE? Perhaps we should consider including a definition of Cyber-Physical Environments?

8. Pardon me for living too close to the ground where the IoT is concerned and therefore getting into too much detail but I (please note this is a personal opinion) think that no introductory (i.e. non-technical) document on the IoT will ever be written without a mention of LogMeIn's Xively (formerly Cosm formerly Pachube). Okay, so this might be irrelevant for 'this' document. It's just that I can see an explanation for how the "Standards" which are mentioned in the document were selected but I am not sure about the criteria being used to select the "companies" that are mentioned and how wide the scope is. Whatever the case, I just thought it would add to the completeness of the document if  this group's (Pachube's) contribution is given a mention.

9. Are we interested in a discussion on sub-IoT's e.g. Internet of Animals, Internet of Signs, etc. for this work? Perhaps we want to mention them somewhere if it is not out of scope.

10. Not sure I can see anything on the work of Weiser M, et al. in the document. Okay, theirs may not be THE one important factor to be mentioned when discussing the IoT but when we look at "Ubiquitous Computing" (which contributes to the IoT concepts of EVERYTIME, EVERYWHERE) and when we remember IoT aids "indistinguishability, invisibility, blending in" of technology (which are some of what Weiser meant when he wrote the work on the "Computer for the 21st Century"), we may wish to consider giving the work a mention - just for the sake of "relative completeness". Having said that, I suppose if "IoT Characteristics" are given a section in this work then Weiser's work will be mentioned.

11. I do not have the reference for this (been searching online for it for a while now) but I read somewhere some years ago that the function of the IoT can be summarised in the 3 C's: Communication, Computing and Control. I do not know if this can perhaps be mentioned as part of a section on the applications or aims of the IoT(?)

12. More on definitions: words like Spimes (Bruce Sterling) and Blogjects (Juilan Bleecker) etc. and their definitions probably require a mention. This is just a suggestion; feel free to ignore it! :) .


B. Questions

1. Quick question, what was the criteria used to select the papers, books, websites and other sources that were referenced in this document? Same as with the industries' list - focusing only on big industry players (or in this case, publishers) risks missing out on the rich field of data and research already discussing and tackling IoT-related challenges. At this point I suppose I should declare a personal interest in 'IoT-related publications' having written a few myself. This fact however does not affect the point which I am trying to make: that not ALL of the most useful published materials that I have consulted as part of my research into the IoT and related subjects were published by the big name publishers. Just a suggestion that we may wish to look further afield. 

2. Another question: Is there currently a universal IoT symbol, logo, etc. that we may wish to include in the document?

3. I know this is a draft (and again, a big well done to the team) but -  and this only because it is a growing document and this growth will probably accelerate rapidly now that the document has been made public - perhaps we may want to consider generating it using, for example, a tool like Latex which will give us a "crisper" interface, clickable links, among other features. Only really for the sake of navigation and readability, etc. You probably already have this in mind but, just wanted to mention it.

4. Lastly, do you by any chance have an RFC in mind for this work? ***By the way I have a draft IoT definitions document that I started a while back (it is nowhere near the scope of this work, nor is it meant to be; just a collection of one to five liners). If anyone is keen to work with me on that - let me know. It's just a small project.



Thank you.


 

 

 

 

 

2015-06-17 @ 3:33 PM by Vulpe, Alexandru

Section 2.2 is a nice bit of history, but I think it could be shortened or put in an appendix. 

White papers are mainly from 2008-2011. I am sure there are more recent papers addressing the topic. 

Recent white papers cover also the integration of IoT in existing communication networks see Nokia's white paper (http://4g-portal.com/nokia-lte-m2m-optimizing-lte-for-the-internet-of-things-whitepaper). Also other white papers like EU-China white paper on IoT identification (http://www.miit.gov.cn/n11293472/n11293832/n15214847/n15218338/n16261917.files/n16261914.pdf)

I don't think a sensor operating system listing would be appropriate, seeing as they come and go from active development stage. 

Section 5 is kind of awkward to be at the end. Definitions should be at the beginning, shouldn't they ?

 

2015-06-18 @ 7:35 PM by Pietrzykoski, Anthony

Paragraph 4.3 quotes Wikipedia as stating that MQTT is a TCP/IP protocole, while true, However, there are two versions of MQTT; MQTT and MQTT-SN. Please update the document to reflect this fact. Figure 1 in the specification would be good to include in your document to show how MQTT-SN interfaces into MQTT.


MQTT V3.1.1 is now an OASIS Standard.

Insert into Glossary:
MQTT-SN (MQ Telemetry Transport for Sensor Networks) is a lightweight MQTT that does not require a TCP/IP stack. It can be used on other protocols such as a serial link like Bluetooth, RS-232/422/485, Z-Wave, ZigBee and the like. However, -SN does require a bride to translate MQTT-SN messages into MQTT messages.

Insert into paragraph 4.3 MQ Telemetry Transport (MQTT):
From MQTT For Sensor Networks (MQTT-SN) Protocol Specification:
http://mqtt.org/new/wp-content/uploads/2009/06/MQTT-SN_spec_v1.2.pdf
http://mqtt.org/documentation

 

The architecture of MQTT-SN is shown in Fig. 1. There are three kinds of MQTT-SN components, MQTT-SN clients, MQTT-SN gateways (GW), and MQTT-SN forwarders. MQTT-SN clients connect themselves to a MQTT server via a MQTT-SN GW using the MQTT-SN protocol. A MQTT-SN GW may or may not be integrated with a MQTT server. In case of a stand-alone GW the MQTT protocol is used between the MQTT server and the MQTT-SN GW. Its main function is the translation between MQTT and MQTT-SN. MQTT-SN clients can also access a GW via a forwarder in case the GW is not directly attached to their network. The forwarder simply encapsulates the MQTT-SN frames it receives on the wireless side and forwards them unchanged to the GW; in the opposite direction, it decapsulates the frames it receives from the gateway and sends them to the clients, unchanged too. Depending on how a GW performs the protocol translation between MQTT and MQTT-SN, we can differentiate between two types of GWs, namely transparent and aggregating GWs, see Fig. 2.


They are explained in the following sections of the specification.

2015-06-19 @ 10:51 AM by Rotondi, Domenico

Hi all,
as one of the person in charge of maintaining the paper and address comments and suggestions thanks to all of you for your hints.
As a general rule, I kindly ask the people who as suggested adding or revising sections to provide me a short text as a Word document so that I can integrate it in a revised version ("short" of course can depend on the kind of proposed integration, but as a general suggestion I think it can be between half a page to 1-2 pages) . Of course insert figures if they are necessary for the clarity of the txt.
Thanks in advance.

Let me now try to address some of the comments:
@Weiler: Hi James, with reference to the WWII section let's retain it for the time being. Regarding your suggestions on wireless technologies and issues, I agree they are relevant in many IoT contexts (e.g., WSN), even if they are not specific to IoT. Additionally, the idea of the document was not to cover all issues and overview all technologies.
Personally, I think your suggestion can be better managed as follows:

  • adding a 1 page section in chapter 3 summaring the technical issues you highlighted
  • adding a 2 pages appendix reporting more technical details and references to the relevant specs

Do you agree? If so please provide me the related contributions.

@Dasgupta: Hi Arup, do you have specific suggestions on what to cut in the RFID section? Regarding approaching the topic from the "end-user perspective" I agree in principle, but we are technicians and, so, it is difficult to have a different perspective until end-users are coming in and suggesting how to improve the contribution adding more user-oriented perspectives.

To be practical what I suggest is:

  • to "advertise" this initiative among end-user communities soliciting comments and contributions
  • to enhance chapter 2 with links to other projects/initiatives so to provide access to actual end-users services or experiences
@Cardalliaguet: Hi Diego, I look forward to get your integrations about Microsoft, IBM, etc.
For sure we cannot exhaust all players in the field. So the open question:
which industrial players do we have to mention and why?

@Oriwoh: Hi Edewede, thanks for the long list of comments. As you stated, this is a start and the main objective of the document was to foster discussion and the identification of the most relevant and critical aspects, as well as of related technologies and appraoches.
I personally don't think the current approach of providing a PDF document is the most effective as compared, for example, to set up a wiki where to organise the different topics and allow the "document" to grow and expand as necessary.

Coming to your suggestions:
  1. yes the legal aspects are relevant. Unfortunately the authors of the draft were not expert in this field and that's why it is not covered. So any concrete contribution to cover these aspects is welcomed. BTW: I think the legal issues tied to IoT are not currently being adequately covered anywhere, but for sure there will be the non trivial problem of providing an overview of how, and if, these issues are being discussed and addressed in the different Countries
  2. this, in my view, is related to the previous point; isn't it?
  3. see answer @Cardalliaguet
  4. not clear to me if you think that the IEEE definition of IoT is too narrow, or if the overview in the doc about the IEEE is to short. In the former case we can only forward the issue to the related IEEE groups; in the latter case please suggest how to expand the overview on the IEEE activities
  5. I personally don't agree we have to add a "definition of Big Data", while I agree that we have to mention Big data, as AI and distributed computing/reasoning, in and for IoT. Any concrete suggestion is welcomed
  6. not clear to me
  7. Section 2.3.6 is summarising the NIST approach/activities. As evdient from other sections the authors are not equating CPS=IoT
  8. this suggestion is related to the 3rd one. Xively for sure has been a relevant promoter (as others, e.g. IFTT). See my comment to your previous suggestion
  9. let's see what are the opinions of the others
  10. please suggest where to mention and how to mention the Weiser work
  11. do you think is it necessary?
  12. please provide a contribution

Now you questions:

  1. there was not a scientific or objective criteria. The authors mentioned the papers/books/... they were aware of and they think were more relevant. Of course thinking to be exhaustive is illusory. A point in favour of the authors is that they have not mentioned any of their papers -:) . Improvement of the references are welcomed
  2. not to my knowledge. If someone is aware of a symbol which is largely used to characterize IoT and that can be freely used, please propose it
  3. The issue you raise is well known. Personally I don't use Latex; as a wrote above perhaps a wiki is a better solution
  4. do you think we have to produce an RFC to summarize this work? Perhaps the net effect will be to have just another document to be taken into account!

@Vulpe: Hi Alexandru, I'll take into account the suggestion to cut (or drop?) section 2.2.

Regarding the rfeferences, see my answer to @Oriwoh and, please, provide some additional relevant ones.

A survey of IoT related OS is, in my view, useful. Perhaps we have to add additional OSs there; what do all of you think?

The rationale of having Section 5 at the end was to try to suggest IoT characterizing elements after having provided a survey of the "state of the art" not to "define something" to be used while reading the document. As stated the idea was to stimulate the discussion and foster the identification of common views, priorities, ... So if most of you suggest to move that section at the beginning we'll do that.

@Pietrzykoski: Hi Anthony, I'll add the MQTT-SN definition in the glossary. With reference to the integration of section 4.3 with the text you provided, I'll do that if others don't disagree. Anyway, take into account that that there are other pub/sub protocols (e.g., AMQP, DDS) that are relevant but are not mentioned

2015-06-19 @ 5:06 PM by ORIWOH, EDEWEDE

Hello Rotondi, Domenico

Well done looking at the feedback.

I suppose a wiki will do it but “where will it end” is the only question especially if the aim is to produce a 'finished' document after a while?

 

1. You're right about this point - countries (USA, EU) are gradually bringing in policies and laws such as the EU's requirement for drones to be regsitered (was on the news the other day); not flying drones in certain areas, 6-month faulty tech cover (was also on the news the other day), etc. However a bigger discussion needs to be had in this regard.

2. Yes, to an extent it is!

3. Okay, okay. I suppose the list just needs to be expanded: Cisco, Samsung, Car manufacturers, Contiki

4. The IEEE definition of IoT is too narrow. A moot point since it's already been published but just thought I’d mention it.

6. (This is similar to the question in 5): i.e. do we need to define terms like Big Data, Activities of Daily Living (health is a big part of IoT discussions), smart homes, etc. or are they outside the scope of this work?

10. IoT beginnings because he and his team coined the phrase "Ubiquitous Computing". I just think that his work and the work on Pervasive Computing are sub-sector forerunners of the IoT just like miniaturization, sensor technology, wireless sensor networks, etc.
.

11. I think it is - but I have no reference for it so we may have to leave it out.

Spimes
12. a.): https://mitpress.mit.edu/books/shaping-things

12. b.) P. McFedries, "The Age of Spimes [Technically Speaking]," Spectrum, IEEE, vol. 47, pp. 25-25, 2010.

My own definition in an article:  A Spime "is formed by a combination of the two words ‘SP’ace and t‘IME’. This word was coined by Bruce Sterling and represent objects that can be tracked and are traceable." Please feel free to rewrite to avoid referencing.

Blogjects
12. c.) : N. Nova and J. Bleecker, "Blogjects and the new ecology of things," in Lift06 Workshop Http://tecfa. Unige. Ch/~ nova/blogject-lift06. Pdf, 2006.
12. d.) http://archleague.org/main/wp-content/uploads/2009/10/SitTech5_111109spreads.pdf

 

My own definition: "Objects that are able to communicate their status by blogging."

 

Questions

4 If the IETF’ll accept it, why not? It will not be another document to be taken into account; it will be the “IEEE-led IoT Reference document”. However, if the gravitas of the document as it stands (IEEE-only publication) is enough without an RFC then okay, we can just go ahead as is. 

 

Thank you.

2015-06-19 @ 6:46 PM by Barros, Tiago

Hi All,
first I would like to say thanks to all of you that created and contributed to this document. It is becoming quite useful in our research.

One point that maybe we could add in section 5 is about the plethora of platforms, protocols and communication technologies that we have nowadays in the IoT field. A question that I hear a lot when I give talks is: what will be "the protocol" for the internet of things?
A major problem is exactly the lack of a standard for wireless communication technologies and protocols in the IoT field. We have "multiple standards", as it was described in this document, being adopted by many companies an the academia, which means we do not have "a standard".
My opinion to this question is that we will not have one "single standard" for Internet of Things simply because each existing standard aims to address some specific application needs. And what will be presented to the end user, and what really matters when we have 50 billion connected things, are the applications and the value they add.

So also a good "definition to the Internet of Things" is similar to the phrase that I heard from Bill Morelli, IHS Director, that says the following:

"The important thing to remember about internet of things is that it's a conceptual framework. It’s not a specific technology, it’s nor a specific device.
It’s really about this philosophy of being able to embed the connectivity and share data across multiple devices, and the value that can bring."

I believe that our challenge in building the infrastructure that will enable this internet of things, is to put together the several wireless communication technologies and protocols (in the physical, data link, network and transport layers) so that they speak a "common language" in the application layer. And in that layer, the things, made by several manufacturers and using these several wireless technologies and protocols, can "talk to each other".

Maybe we could also add to section 3, a view that is becoming true: although we have this diversity of technologies regarding IoT protocols and wireless technologies, one thing is also becoming a de facto standard: the architectural component view of the internet of things. Observing many IoT platforms, we realize that we have an architectural pattern forming: the things talk to a gateway to reach the cloud and the applications access the cloud to send and receive data to the things.
An evolution for this architecture, will be when the things have sufficient processing power and connectivity at a reasonable price (the price today is still high to embed an IP into each lamp of the world), to access the cloud by themselves. They will have an IP(v6) and incorporate the gateway functionalities, and the gateway will disapear.
And the Architecture 3.0 will happen when, again by the evolution of processing power, these things become autonomous, accessing each other's data and taking decisions based on complex AI algorithms.
These architectures are not exactly mutually exclusive, they will coexist, according to the application's needs. However, they will happen in that order of evolution of processing power and communication.

Well, here are my first contributions to this document and, if you agree, I can prepare a doc with some pages to address these two points and discuss more about that.

Thanks again,
Tiago.

--
Tiago Barros, Sr. Engineer
Head of CESAR IoT Research Lab
Http://www.cesar.org.br/

2015-06-24 @ 3:37 PM by Dasgupta, Arup

@ Rtondi, Domenico,

"Hi Arup, do you have specific suggestions on what to cut in the RFID section? Regarding approaching the topic from the "end-user perspective" I agree in principle, but we are technicians and, so, it is difficult to have a different perspective until end-users are coming in and suggesting how to improve the contribution adding more user-oriented perspectives.

To be practical what I suggest is:

  • to "advertise" this initiative among end-user communities soliciting comments and contributions
  • to enhance chapter 2 with links to other projects/initiatives so to provide access to actual end-users services or experiences"
To get a user perspective why not tie up with IEEE SSIT and the Humanitarian Activities and look at their activities? 

Re the RFID section there is a whole lot of introductory material which is of historical importance but could be trimmed to make the section more readable.

2015-06-25 @ 3:00 PM by Dasgupta, Arup

If Section 2.0 is State of the Art and 2.1 is Introduction then 2.2 should cover ALL items related to IoT and not just RFID. My suggestion is that the Historical Background should be an Appendix and not in the main body of the document.

UN-GGIM is also supporting th IoT for geospatial applications. Please see document on Future Trends in Geospatial information Management: The five to ten year vision. The following is an extract from this document
"1.3 Linked data and the ‘Internet of Things’
1.3.1 Given the vast amount of data being generated, particularly through use of the Web, and the need to make sense of this data, the ability to link information on the Web will be increasingly important in the coming years. To this end, we may see data increasingly being distributed as ‘linked data’ in the coming five to ten years. Linked data offers the opportunity to connect data to other pieces of data on the Web, contextualising and adding value to the information that already exists.
1.3.2 Semantic technologies will play an important role over the next five to ten years when it comes to publishing and making sense of this data, offering the opportunity to create rich machine-processable descriptions of data. This will enable knowledge sharing and reuse in addition to data sharing and reuse. It is expected that data will really start to show its true values when it is combined with other data sources. Location will provide a key underpinning framework to the Web of linked data, providing an essential information hub that brings many datasets together.
1.3.3 The network of tomorrow, built on an increasing numbers of sensors and thus, increasing data, will produce a hyper- connected environment or ‘Internet of Things’, with estimates of over 50 billion things connected by 2020. The ‘omnipresence’ of geospatial information in our lives, whereby almost all pieces of data have some form of location reference, will continue, with location providing a vital link between the sensors that will generate the Internet of Things and the Uniform Resource Identifier (URI) assigned to a thing or object within that connected world of things. This will drive the demand for informative standardised metadata as part of geospatial data, in order to maximise usability.
1.3.4 We are increasingly likely to see geospatial information in demand to assist the evolution of this connected ecosystem over the next five to ten years. The emergence and use of precise location information in this way offers great opportunities and will see it form a core part of information technology infrastructure. Nevertheless, use in this way will also present geospatial management challenges over the coming years. ”

I also have an article at: http://geospatialworld.net/uploads/magazine/January-2015-Geospatial-World-Magazine/files/78.html where I have covered some of the applications aspects of IoT

2015-07-03 @ 6:53 PM by Rotondi, Domenico

Hi all, thanks for the intersting discussion. I think one of the main objectives (foster the discussion) of the document is being achieved.

It is becoming difficult to try to summarize all points in a single document as done at the beginning, both because I need more structured contributions (otherwise we risk to have a new version but with many placehodlers or incomplete sections) and becuase the raised issues are becoming too many.

I'm still waiting to have hints on possible solutions like a wiki where registered people can directly integrate or amend the document.

As another general comment, we have to take into account that any document cannot be exhaustive being the IoT scenario so wide and with so many actors and new initaitves that cannot be completely tackled.

So let's try to use the discussion and this page as:

  • a place where to try to identified the most relevant and critical aspects (especially from the point of view of the IEEE and its IoT Initiative)
  • a place useful to circulate news and information about IoT so to give to all of us the opportunity to be up to date
Let me come to specific points.
@Tiago:
regarding the plethora of protocols, platforms, etc. I don't think there will be 1 (or a few) platform(s) and a reduced set of protocols. IoT covers so many domains and will have so many impacts on our lives that I don't think single solutions could be effective in addressing the needs of all these application domains. I personally thing that the effort has to be:
  • assuring a few, suitable (i.e., optimized for WSN, suitable for PANs, for HANs, ...) protocols be defined for the lowest layers (i.e. layers 1 and 2). I think this is the objective of the IEEE P2413 and 802.xx WGs
  • having layers 3 and 4 based on IPv6 and TCP/UDP so to assure that, potentially, any IoT device can communciate with any other device or service
  • the upper layers using more domain specific protocols (hopefully based on a restricted set of application transport protocols like HTTP) but with enough meta-information to assure bridges or gateways can be easily (and hopefully automatically) developed and deployed (to this end it seems to me the AllJoyn Device System Bridge is a good example).

The big issue, from my point of view, is not on the lower layers (where I'm expecting there will be a well defined set off protocols addressing specific communciation needs), but the "interoperability" at the higher layers. This will be furhter complicated by the change in the "interaction patterns" among entities (e.g., devices and applications). Nowadays, these patterns are more or less stable in time and pre-defined (a device communicates with other devices/services as defined by the device/service owners or administrators); while in the IoT world these patterns will evolve to short-lived, often casual or spontaneous ones. This implies a huge change in the way we will develop devices and services (e.g., moving to event based and proactive approaches), as well as in the way we address issues like trust, identity managment, authorization, etc. For example current approaches and technologies for identiy management and authorization doesn't scale as IoT requires.

I don't agree with a view centred around the cloud where devices, at the end, are just able to generate some data/events and communicate with the logic that is on the cloud. My smart phone/tablet is smart enough (and will become even smarter) to locally manage a lot of things without necessarily have to use the cloud (OK, I know my Android device uses a lot of Google cloud services, but this is a "marketing" choice made by the Android OS developer!). I'm expecting my house will have a hub smart enough to manage all data/events I want to remain local and only provide to the outside data I want to go out of my house and my direct control. Smart cars, as all devices/contexts (e.g., factories) where safety is critical, will be able to perform locally all required processing.

Fog computing promoted by Cisco (or Edge Computing by IBM), as well as user-centred approaches (e.g., Kantara UMA) are moving along this line.

To come to your last point, I'll appreciate some structured pages on these aspects as you proposed. Thanks

@Arup:

sorry I don't konw the "IEEE SSIT and the Humanitarian Activities"; so if you think they are relevant, please provide some specific contribution.
Regarding §2 I'll move some part of it to an appendix.
I think you are suggesting to include in §2 the statements you report on UN-GGIM, isn't it?

2015-07-30 @ 8:34 AM by Chaudhuri, Abhik

First of all I would like to thank IEEE IoT Community for taking the initiative to properly define Internet of Things. This is one of the key requirements for enabling the future of 'Things'.

After going through the document I feel that the 'trust' aspect should be included in the definition considering all scenarios (small/large environments). The reason being, we need trust based service models for all kinds of IoT implementations and trust factor encapsulates the other binded factors like security and privacy.


Regards,


Abhik Chaudhuri

Chevening TCS Fellow in Cybersecurity & Cyber Policy

Domain Consultant: Cyber Security - IoT and Smart Cities

Tata Consultancy Services.

 


 
 

2015-08-04 @ 5:34 AM by Kumar, Arvind

This is a very comprehensive document. Congratulations to the Roberto,Abyi Biru and Domenico for this initiative and taking the IoT platform to the next level.

As rightly mentioned, security layers are very important to protect individual rights and aligning with prevelant security standards of respective countries. RFID evolution to the next level is crucial. Communication protocols upgrading is key to success of this technology. Implants within human body needs a new look forward. And of course, a comprehensive definintion is the key to spawn development of IoT.
Segmentation will become seamless as rigid boundaries will melt away. While segmentation is a method to make things manageable, ultimately the platform will look more like fuzzy interactive boundryless entity weaving around the biggest beneficiary that is we the humans. 
Looking forward to a challenging drive for IoT to evolve and develop as the next revolution in human history.

Arvind

arvind.kumar@ieee.org

2015-10-25 @ 1:22 PM by Haller, Stephan

I value the document, as it summarizes well the activities in the IoT space, and I think it also provides a good start for someone new to IoT. Two things though I find unfortunate:

  1. Chapter 3.3 talks about addressing, but it mixes two slightly different concepts: An address and an identifier. These may be the same, but that is not always the case, hence I think it is advisable to make a distinction between the two. An IPv6 number is an address that I need if I want to communicate to the "thing", while an EPC is merely an identifier for a thing (e.g., in a database), but it doesn't say anything about what address to use if I want to communicate with the thing. Clearly, an IPv6 number can be used both as an address as well as an identifier, but not all things in the IoT have an IPv6 address. In addition, the same physical entity (or actually - in IoT-A terms - the virtual entity representing it) may have multiple identifiers, e.g., a general identifier, company-internal number, a domain-specific number (SGTIN, SSCC, GRAI, etc.).
  2. In my opinion, the not all "Things" are directly connected to the Internet, but they still qualify as a "Thing". IoT-A took that approach when defining physical entities. A physical entity relevant for IoT is also called an "entity of interest", and it thus may be observed by sensors in its environment. Take for example animate physical entities like humans or cows. They are hardly connected directly - a cow may have an RFID chip embedded, but most humans would not agree to being tagged - but they can be tracked using presence sensors, cameras etc. The "Thing" or entity of interest is still the cow or the human and not the chip or presence sensor.

2016-02-08 @ 5:25 AM by Maram, Arvind Reddy

We at yzThings perceive IoT as "The application of science and technology to make our world more interconnected, productive and eco-friendly!"

2016-09-14 @ 3:57 PM by Perez, Felix Armando

It is a very useful document.

Since 2012, I am interested on all about IoT; here in UNMSM we have created a introductory lab of IoT, providing some insights and training to undergraduate students.

Specially I am interested in the management of large IoT systems using autonomic computing but I see only two words about in this first revision, please anybody could provide more insights about?, thanks