Hi Robin,

Concerning your question of “islands of applications with no interoperability” let me share my thoughts with you.

Firstly I’d like to distinguish the service platforms roughly into two parts:
1. Platforms for MNO’s for the operation of M2M services
2. Platforms and systems for integrating and operating M2M applications.

Whilst the first group is addressing the MNO’s with many standardized systems I do not see the danger of interoperability issues there. Of course there is a lot of adaption, customizing or whatever is necessary to get the system up and running, but the roads are paved in some way.

With the second category things are different. As you have mentioned some companies are coming from the hardware side and approaching this issue with support of standardized protocols. To this I would count to some extend also the IP services integrated in the current GSM module families. For example the e-mail interface in the TC65i from cinteron would fit to this category. Other hardware manufacturers do a higher integration and design for example a HTTP based protocol set.
Looking from the other side there are many vertical solutions that try to expand their functionality. Whilst for example tracking and tracing solutions in the past were focusing on location services new generation of these systems also transmit information available on the CAN bus such as speed and mileage information or even temperature information for the fright. However these applications have a fixed, web based layout and integration has not been a focus issue with the system design.

For these two types of applications I do see a danger of “lost islands”.

However with a new generation of M2M integration software (I’d like to call this software thingware as this software is intended to connect the internet of things with IT systems) the system design focus has been to integrate scattered devices into complex and diverse IT systems. An example of such a system is x-active (http://www.ettex.de). Analyzing a M2M system the interfaces to the different hardware (such as vending machines, forklifts, cars) are uncountable, the processing power of a device is limited and the cost of wireless information transfer is relatively high. Furthermore a complex IT landscape can be found at potential M2M customers. As a consequence interfaces to the hardware on the devices must be versatile, the traffic over the air must be kept low and the interfaces to the IT systems must be flexible and adaptable. It is advantageous to use a concept of adaptation instead of hard coding for integration – this idea is comparable to the SAP concept of adaption (and no coding) to the customer needs.

To me this concept of thingware is the future with the integration build in by design.

Do you agree with my thesis?

I’d like to comment Robin’s and Thomas’ statements. It is true that there are a lot of ‘platform’ providers already in the market and for sure we’ll see even more of them the next years. I agree with Thomas that we clearly have to distinguish between platforms for MNO’s, which are purly SIM oriented and platforms for solution providers and customers. Most of the platforms I’ve seen in the market are verticalized, meaning they focus on specific market segments. The thing here is, if I want to use it also for other segements, there is a lot of work to do in order to adapt it to the specific requirements in that segment.
Wouldn’t it be much better to work with a horizontal platform, which is independent from vertical segments? A platform which provides all necessary communication layers in a stadardized way and offers open interfaces for the vertical applications sitting on top of it?
Looking at the core competences of customers and solution providers, it’s very clear, that this is not the communication part, but the specific vertical part. Consequently they need a platform which can handle all communication staff and deliver the data from the field which they need for their business model.
The only horizontal platform I know is x-active from ettex, which is also market by M2M Simplexx.
Look at it and let me know your thoughts about it.
http://www.m2msimplexx.de/index.php?option=com_content&view=article&id=122&Itemid=66&lang=de

Ulrich Kaindl

Hi Thomas

Thanks for your thoughtful comments. You distinguish M2M platforms into two parts – basically platforms for MNOs and then for applications. While I agree that there are two levels of M2M platforms, in fact MNOs are interested in both levels. The most obvious example of that of course is AT&T’s current partnerships with Axeda, ILS Technology, SensorLogic and Sierra Wireless in addition to its longer-held partnership with Jasper Wireless.

I think there are potential problems on interoperability at both levels. With so many platforms now reaching the market, it is inevitable that the integration work required for each platform to interoperate with any other will not be followed through. Choices will be made by platform providers to integrate with some others, but not the whole universe. This may be especially the case for the application part, as you put it, but it is a wider issue than that I think. Although standards will help to an extent, it does depend on those standards first of all being there, and then being adopted.

I like your reference to tracking and tracing solutions now looking to transmit other data to expand on their service. My thinking is different but related. In particular, I have in mind three practical levels where interoperability will be needed. Firstly, a company using one platform that acquires another using a different one and needs to share data on the two different platforms to provide a new, integrated service. Secondly, a company using one platform that wishes to partner with another using a different one in order to provide a new, integrated service. Both of these examples will typically involve companies operating in the same or closely related business sectors. Thirdly, though, companies wishing to utilize data from applications outside their current sectors to create new service opportunities. Each of these can of course be resolved as individual projects if enough money is thrown at them. The higher the cost, though, the less likely they are to be implemented and that means new and innovative service ideas being stifled.

These go together with an underlying concept of the Internet of Things as being able to share data across application boundaries to create new service opportunities. However, this is not saying that the new platforms in the market are too vertically oriented. I don’t think they are – see my other comment to Ulrich. It is more about the straight interoperability. Certainly, your idea of thingware helps to address the issues of diversity on one platform – but across platforms? Not sure.

Hi Ulrich

Thanks for your comments. I have partially responded to these with my reply to Thomas. Some other points I would like to address though.

As we all agree, M2M has grown up in vertical silos where each individual (vertical) user application has been developed expressly for the purpose of getting that one application to market. As a result, application developers have “re-invented the wheel” relentlessly and at great expense in terms of costs and resources. It is a key reason why M2M solutions have appeared to be complex, costly and often taken a long time to get to market.

However, you should distinguish between vertical M2M solutions (of which there are many) and more generalized M2M platforms. Attached to my blog last week is a list of 46 platforms in the M2M market. This is not all of them (x-active is another) but what they have in common is that they are all horizontal. Some are applied to a particular sector, but in principle all could be applied to any sector.

Any thoughts on that?