Home Automation and IoT

Home Automation and IoT

Home Automation and IoT

Home Automation as a Technology

The primary objective of technology is to improve lives. When I say lives, I of course do not limit it to lives of human beings, but all life forms that exist on earth and the earth itself. This is amply evident from the fact that the large technology companies are shifting their focus from consumer (which is primarily human beings) centric to a more socially and environmentally responsible view. Sustainable development is the need of the hour and it is expected that any emerging technology of the day would touch human lives in a responsible manner.
This should be the guiding principle behind any home automation offering. Obviously such a solution must ease the life of the consumer deploying it, and we’ll discuss in detail about few of the numerous ways that home automation would enhance one’s life. However that should not be the sole focus of such a solution. It should also strive to make the earth a better place to live. It should not only turn a home into a Smart Home, it should also turn it into a Green Home.

What is IoT

If you are “connected”, it is highly likely that you have heard of the thingy; the Internet of Things or IoT in its short and sweet form.  IoT is essentially a collection of “thing”s connected together with (I)nternet (P)rotocol. These “things” theoretically can be anything; it’s just left to imagination. Though normally IoT is considered to be Machine-to-Machine (M2M), they can theoretically be any object, man made or natural as long they can be addressed using IP and possibly be fitted with sensors and/or actuators.

So, IoT essentially is a framework to interconnect objects that may serve a common purpose when connected together. These connected devices communicate with other to exchange information which essentially generates data. The data is processed at various levels and decision is made based on the data at different places in different ways.

For example, connected cars can exchange various attributes like average speed on a road, waiting time at various junctions for your in-car navigation system to compute a more efficient route. Or integrated sensors can monitor the health of important installations like bridges and dams in real time and produce data that can potentially save large number of lives. Wearable devices has given rise to the potential of attending to emergency health conditions without delay. The possibilities are just infinite. So the conclusion is IoT can touch human lives in much more real sense than most of us think.

The following diagram from IoT World Forum is a representation of the IoT model. Like the OSI seven layers of networking, we have a seven layer model here too.

At the edge or Layer-1 of this model, lies the “things”, the sensors, the actuators, and other physical devices carrying out the task of sensing and controlling. These devices are connected with each other and they communicate using some communication protocol. Layer-2 deal with the connectivity between these devices.

The communication between these devices generate a lot of data. This data has to be captured and processed to generate intelligence out of this. The layer-3 of this model deals with processing this raw data and converting this unstructured data into structured data that is then stored in layer-4 of this model.

Layer 5 provides abstraction of this large amount of that and present the data in an aggregate form to the applications that are interested in this data. The applications form the layer-6 of this model. The applications are the components that derive intelligence out of the data by analyzing them and present the data in a form of reports that are understandable to the intended users. They can also make decisions and control.

Finally, the most important layer of this model is collaboration. This is the layer where users – human beings and business processes alike – interfaces with the system. The users are at the center of the model. Reports are presented and control provided to the users here. Secured access to the intended users is the most important aspect of this model.

Security encompasses all the layers of this model. Access control, roles, and privileges are enforced at every layer of this model. Starting from secured communication at layer-2 to secured data access to finally the access control at the collaboration layer; security is central to reliable IoT system.

IoT and IPv6

As we know by now, at the core of IoT, lies a number of objects connected using IP. This connectivity is crucial to the philosophy behind IoT, real time communication. The primary bottleneck behind connecting a large number of devices to the internet is exhaustion of IP (IPv4) address space. This limitation is even more prominent in a developing nation like India, While Network Address Translation or NAT serves as a workaround, it has many limitations; the inability to directly connect to a device with NATted IP address being the most prominent one.
IPv6 was conceived to solve many IPv4 limitation, the address space limitation being one among them. By increasing the address field by four folds, from 32 bits for IPv4 to 128 bits, IPv6 increases the address space from 6 billion to some 340 trillion, trillion, trillion addresses. Which means enough address to connect any object you may wish. This means Internet of Things becomes a reality with IPv6.
The other thing is about being green. There is a concerted effort in standard making bodies like IETF to standardize low power sensor networks over IPv6 like 6LoWPAN (IPv6 over 802.15.4). The low power sensor networks are the very basis of what forms the IoT cloud. So IPv6 is going to play a central role in the success of IoT.

IoT and Big Data

Let’s first understand what is big data and what is the correlation between big data and IoT. As the name implies, big data is large set of data. Traditionally most of this data has been generated by us, through our clicks, through our typing, through our searches, uploads, posts and so on. Every digital activity can be a source of data. It can be easily understood that when the source of data is so large in number, the volume of the data, the velocity with which it gets generated, and the variety of the data is going to very high. So is going to be variability. On the other hand, the veracity of the data can be questionable at times.

Now, though we understand the generation of big data, do we really care about most of it? Of what value would this data be to anyone anyway? However random and non-useful this data may appear, it contains large wealth of information that when analyzed properly. They for instance, can help Google to place ads that are relevant to you or may help provide with browsing suggestions. This data can be so useful, they may even lead to rise and fall of governments as was speculated in the last Loksabha election of India!

Now let’s understand the role big data plays in IoT. As we understand by now, the layers 3 and 4 of the IoT reference model deals with data. Now the amount of data that has to be handled at these two layers depends on the number of devices handled at layer 1, and the kind of communication used at layer-2. An IoT deployment may deal with huge number of devices each contributing to a large amount of data frequently. So the amount of data that may have to be handled at layer-3 and 4 can be very large. Traditional computing and storage systems would prove to be inadequate to handle such large amount moving data. One needs to use big data technologies for processing such large amount of data.

Internet vs. Intranet of Things

We have discussed about connecting a large number of devices to the Internet and how IPv6 would play an enabler here. With IPv6, we’ll be able to connect to Internet each and every device that we wish. While this is great and it opens up a lot of opportunities in the space of IoT, we need to understand if  we really need to connect each and every device to the Internet. Connecting a device to the Internet, among other issues, comes with the inherent risk of unauthorized access. A device generating sensitive data or carrying out some critical task, if not properly secured or is vulnerable to some form of attack, may be a disaster in waiting.
Despite all the advancement in cryptography and related technologies, isolation still remains the most effective form of security. So a very effective way prevent unauthorized access by large number of hackers out there, is to isolate the devices that do not really need to be connected to the Internet. Access in any form, should be purely on a need basis and controlled. So while designing any system, it’s of paramount importance to segregate devices based on the required level of access to be given to them, designing appropriate layers of security and access control for each group.

So a hierarchical access model may suit most use cases. In the IoT reference model, most use cases would require the “center” to be more externally accessible and thus would most likely be connected to the Internet and possibly would be accessible with other methods of access as well. Also it may be possible a few of the lower layers might be hosted in a public cloud and thus would be accessible over the Internet. All these layers should implement highly secured interfaces to prevent unauthorized access. However, the deeper you go down the layers, the lesser you want them accessible over a public infrastructure.

At the “edge”, you have a large number of devices or “things”, which are doing the actual job of sensing and controlling. These devices, if exposed to unauthorized access, can cause serious damage. Implementing appropriate access control for these large number of devices is not an easy task to accomplish. Also in most cases, you would not really need to connect these to the Internet. So in most cases, it’s the “thing”s that are most unlikely to be connected to the Internet.

Now, does that not against the very fundamental concept of IoT? We did start by stating that IoT comes as a result of the need to connect a large number of devices so that we can generate intelligence out of those. However interconnecting these devices need not have to be over a public infrastructure or the Internet. One can interconnect the “thing”s over a secure private network. These devices can communicate with each other over this private network. This private network can be termed as the Intranet of Things.

However they still need to communicate with the upper layers. So what is required here is an interface and proper abstraction for the upper layers to communicate. The upper layers will communicate with the Intranet of Things using the primitives designed for such communication.

Big Data vs. Small Data

Till now, we discussed the relationship between IoT and big data. The main factors contributing to big data were
  • Large number of “things”
  • The “things” generating data at a fast rate.
However, one or none of of the above two assumptions may hold good in several IoT deployments. For example, if we take home automation as a use case for IoT, in most cases none of these tweo assumptions will hold true. You may not be having a huge number of “things” at your home, neither they may be generating a lot of data. In such cases, one need not deploy big data technologies to process and analyze the data. However the IoT reference model is still applicable, only the amount of data needing processing is different in this case.

IoT and Home Automation – Tying it all Together

Let’s now examine how does a modern day home automation fits into the IoT reference model.
Here we are using the working principle of RDHome Gateway – IVR as an example to analyze this.

This solution uses an Intranet of Things model to connect the things. This is a secure in-home cloud that encompasses both layer-1 and layer-2 of the reference model. The devices monitor, sense, and control the home appliances and they communicate with each other and the RD Home Controller over a secure, private, wifi network. This network uses WPA security, on top of that the communication happens over SSL ensuring that the communication is secure.

The RD Home Controller encompasses all the layer from layer-3 to layer-6 of the reference model. It employs a fast CPU with enough RAM to provide it enough horse power to run an advance suite of analytics and control software. The layer-7 here is essentially the house owner who has deployed the solution. A large number of access interfaces are provided, web-based access and IVR based access are the two most convenient access options when away from home. The web interface again uses ssl and user authentication to prevent unauthorized access. The IVR allows calls only from registered numbers coupled with a PIN to prevent unauthorized access.

(C) RD Home Control, 2016

Leave a Reply