Thanks to my current role I have the opportunity to talk to different players in the IoT Value chain. Although I am more focused in the Telecom area, this market brings a broader vision, by monitoring the pulse of this disruptive, fast market.
Things I’ve been realizing about the different players: Everyone wants it, everyone pursue it, but nowadays there is no clear business case on the table. There are too many players, different costs, different business models (and surprisingly old ones…), and in the end, only one service to deliver…
Depending on the player you talk to, he/she sees the challenge on different sides of the value chain, mainly in HW and protocols than in business. That is why everyone is focusing, business wise, on the M2M (approach) uses cases (Home Automation, eHealth…). I mean M2M approach because most of the final solutions delivered to the market nowadays are vertical, that is, own HW, own protocol, own back office, own service, not though to work with other “verticals”, other uses cases, other companies, they are not prepared to be cross business.
Device side of the value chain
Devices costs are falling, but they can fall much more. The IoT market is well known as a collaborative one, each side of the value chain adds its own value and cost, so it is very important to have this market vision in mind when a sensor, a chip, or HW in general, is developed for an IoT purpose, because a sensor alone is not IoT, it is just a HW more or less intelligent, and it is needed to add to different layers more of value (and costs) to reach the business. As a result, device manufacturers have to rethink the way they produce their goods, change the rules, and look for other ways to get the profit expected.
For example, thanks to the way sensors work, devices (included smartphones when they work as a Sensor) don’t need to send huge amounts of data, the payloads sent are very tiny, which is why a lot of Telco Operators in EEUU and Europe are thinking about how to use their existing 2G networks, by not shutting them down, taking advantage of their geographical coverage (close to 100%), and the bandwidth it can provide, enough for the payload size. Nevertheless, the real problem is in the 2G chip cost that hinders this step. This should be closer to 1$ and now is closer to 10$. Why can´t an old technology reduce its cost? Maybe one of the biggest issues is the business model.
Other risks that devices, sensors and end points manufacturers have to be aware of, are the upcomingplayers in their market, that is, Raspberry Pi, Arduino, etc. With these, you can design, build, and launch any kind of little sensors, with your own features, which are flexible, and cheaper. How can we beat this kind of technology in the sensor market? One month ago, a colleague told me his fifteen year old son had developed a connected TV platform he could enjoy everywhere, in every platform, with a Raspberry Pi… so, Intel, Qualcomm, etc have to be aware of their business, in the same way Telcos have to be aware of new service players in their markets (refer to my last post Telco Operators in the Digital Era).
And talking about OTTs, what is the role wearable technology is going to play in IoT…? Now wearables are managing small data, that is, personal data, and most of them are difficult to connect to others, to your IoT ecosystem. Nevertheless, this is moving quickly from “me” to the “world”. Probably this is the area that will decide what is going to be “standard”, let´s see what google, apple, Thalmic Labs, MC10… do in the following months…
Protocol and Transmission side of the value chain
When enterprises think about IoT, depending on their market, they all think about their own protocol stack, and the ways to address an IoT approach to try to take advantage of new services, cost reduction, self-control, performance increase, etc.
If we zoom from the device point of view, and zoom, for example, in a connected home/home automation use case, enterprises start to think how to connect their endpoints, and how to deal with them, Bluetooth? Which version? Zigbee? Is it too expensive? RFID? And if we come back to the previous chapter, what about the chip cost to transmit the payload with the protocol decided? Can it be reduced more? If it can´t, probably, new protocols, with new chipsets will arise, so the mess will be increased.
If the final approach is M2M, not IoT, it´s easy, the solution provider selects the short range protocol, and solves the communication, but is an M2M approach. What happens if the final user wants to use another sensor that uses another protocol, and this is not recognized by the end to end connected home solution? M2M is not a cross business approach.
If we focus on a Telco Operator, their main concerns are about IP Stack, IP addresses management, and IoT network management, as well as radio technologies to be used, that is, 2G, 3G, 4G…? And the future…what is going to happen to 5G? Is it more bandwidth that is needed for IoT? Is it going to be WIFI (and its evolution to Low-Power WIFI 802.11 b/g/n) or WIMAX the final radio technology winner? What are the real features IoT needs in terms of connectivity: Speed, low cost, bandwidth, latency, availability, reliability, coverage…? Is a new radio spectrum needed to cover all these features? New Telco Operators are appearing on the market, just focused on M2M/IoT network, for example in Europe is Sigfox, and could more come up using new MVNO just focused on IoT?
If we focus on utilities, we can see that they are concerned about their CAN communication deployed and how to deal with it… or not. Does the way to manage old and new sensors have to be changed? Do they have to mix different approaches or start with an IoT approach? Are mobile communications already prepared for their needs or try a fixed communication?
But new protocols have come up during the last years, more focused on M2M and IoT environments, taking advantage of the new IP era (OTT approach), like MQTT, AMQP, STOMP, etc. These new protocols “fly” over IP, so the final integration with the current communication infrastructure are easier. But in this field we find the same mess, I have mentioned three different protocols, but there a lot of more, and other strategies, and a lot of different associations with and without Telco Operators racing and competing amongst themselves, in order to be the winner of defining a standard communication protocol. But the reality is IoT is evolving, and is not waiting for the protocol winner…
All this stuff concerns this market, and with the wrong approach, it can increase the cost of the final service to be delivered, in terms of final service price, maintenance, scalability, integration and evolution.
Platform side of the value chain
Platforms are the enabler, the glue that holds together devices, protocols, transmission, and IT infrastructure with final services. Can you develop an IoT service with no platforms? Yes, but your costs and your time to market will increase dramatically, and as I have mentioned before, an IoT service is the addition of values and cost of each layer in the value chain. So from my perspective, if a platform is not considered, probably the value proposition cannot be viable to the market.
This applies to the rest of the layers, as well. Nowadays, there are a lot of platforms in the market that bring the flexibility to connect different kind of sensors, using different kind of protocols, that are able to manage the endpoints, design business logic, easy to integrate, easy to deploy (the easiest way: in the cloud), and that grant companies to build seamlessly new services.
But cost is, again, the point to be solved. There are some platforms that have reduced costs but do not bring the flexibility needed, perhaps useful for personal use, but not for companies. Others are quite complex and heavy, and look more like a typical ERP. Their costs are huge, and this is a problem when deciding what platform you need, depending on if you are a company, individual, or what you need to solve, improve, etc, and the rest of the costs you have to take into account to achieve the services dreamed of.
Another subject regarding costs is the business model. There are a lot of professional platforms that continue to use an old business approach, that is, licenses, and this does not work in this market, mainly, because the market has to trust on IoT, and nobody is going to invest in huge (or not so huge) licensing costs with no clear business expectations.
Let me share with you the fact that in Altran we have solved all of these problems, but do not worry, I will not bother you with adds at this pointJ, but if you are interested, feel free to contact me.
Service side of the value chain
Services, use cases, final value of what customers are waiting to buy. One of the typical questions I usually hear and read is, “what is the value proposition for my business or for me?” Do I need that…? Some use cases just offer nice dashboards with a lot of dynamic graphs, and a lot of sensors that solve zero issues to the business or final customers, and this is causing companies and people to be wary of IoT.
The final use case has to solve issues, to cover needs, to improve businesses, reduce costs, improve the way of living, bring new ways to enrich the communications, etc, with new ways of consuming these services, and, as a result, new ways of paying them. IoT service is not a car, but could be used by a car, my car, but the way this service is bought and used is different.
Many times the focus is more on the previous chapters (devices, protocols…) than on the business to be covered, with no knowledge about it. What is the right price, in which way for X market? What am I going to improve?
It is important to take into account that a well-designed and focused IoT service is going to generate tons of valuable data, to be exploited (so you need to know the market you are focusing on), that can help improve the service as well as reach the final value proposition to the market. This opens the door to another interesting subject, Big Data, but I will leave that for another post.
Cross side of the value chain
There are some subjects that affect almost all tiers in the value chain, and these, have to evolve to reach a maturity level.
Cross Infrastructure side of the value chain
Well, when I write Infrastructure, I mean IT Infrastructure, the base technology that support the final service works. Probably somebody can think that this side does not really belong to the value chain, because it is in every tier, and probably he/she is right, nevertheless it has to be taken into account because it adds value and costs.
As I mentioned in the previous chapter, a platform can be taken into account or not, but one of the features platforms have is to manage this infrastructure, and costs. If a platform is nottaken into account, this management can increase the complexity, and of course the costs (and not the value). So, be aware of this layer, mainly if you are looking for features as scalability, reliability, availability, multi tenancy, in memory computing…
Cross Security side of the value chain
It is a well known fact that security is one of the subjects that everyone is concerned about. Nowadays, traditional approaches are being applied to the security matters, that is, VPNs, Firewalls, encryption, user profiles, etc.
But these approaches don´t seem to be enough due to the implications that some of their uses have on a human being´s life. Imagine that a connected pacemaker was hacked and stopped, or your connected and automated car was not able to brake due to set up modifications, or traffic light management was broken into…
Recently I read an excellent article called “The Internet of Things (IoT) Immune System: Swappable systems with “destroy and replace” architecture” and it let glimpse what kind of security could be applied to the IoT environment. It talks about adaptive systems by using behavioral from the nature, how cells, virus or retro-virus works, how they morph and change, when they are attacked, being able to repel the threat by changing itself by evolving and changing its structure.
Cross Identity side of the value chain
Identity is linked directly to security, but it is more than security. How can I keep my personal data from being used, exploded, or scattered by 3rd parties; how can I control my digital identity?
Since IoT and Big Data techniques are going to meet in the short term, my identity, that is, my habits, my life, will be able to be analyzed, but, does anybody want to be tracked daily? Yes, I agree with you all, today we are all already tracked by Google, Facebook, Linkedin, Whatsapp, even Angry Birds and Candy Crush are analyzing us, this is the price we have to pay for free services and people accept it. IoT is coming to stay, with a broader scope of our lives, so Identity Management is critical and must be taken into account, where the final user has to be able to control his/her own digital identity and footprint.
As you can imagine, the road from performing to delivering an IoT service is more complex than one can expect, but I think it is very important to have a broad vision when a value proposition is done in the IoT market where are a lot of players. However the solution is not just in one layer, but by joining all of them. Now we have the technology, we have the opportunity, and we are living the right moment to have fun with it.