No Jitter is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

UM Meets M2M

The last article in this "M2M in the Cloud" series posed a question: Has it arrived? The answer, as evidenced by short takes on three solutions coming from SensorLogic, Viewbiquity and Palantiri, was a resounding Yes. There are others and of course more will come. Enterprise environments are migrating to the cloud, and putting M2M communications there as well gives IT control of the real-time, actionable data coming from the devices.

IT does not like anything coming onto their turf that they don't control. I'm old enough to recall their objection to PCs. More recently we have seen serious concerns being expressed about the unauthorized use of smartphones, which came into offices via jacket pockets and handbags. That development was covered in my series on the Mobile Enterprise. The hyperlink will take you to the first article.

Different Types of Mobile Data
If M2M solutions are located in the cloud then IT can manage the devices and also enable the integration of device data with mainstream business databases like ERP and CRM. But there has to be a business case.

On the other hand, the business case for the Mobile Enterprise (ME) is predicated on manual access to those databases in order to obtain and input data. This allows management decisions to be based on up-to-date information.

Therefore if M2M integration is viable, then one would expect to see synergistic relationships with ME. And the resulting converged solution would leverage both investments. But before considering the business case for convergence of these two elements, we should look at a fundamental technology issue.

Right now M2M and ME inhabit different domains. Regular IT systems are not set up to handle real-time data: ERP, the most obvious candidate for convergence, gets batch updates from factory floors and warehouses. And most M2M solutions were not designed to store a lot of post-processed data in order to enable analysis at later dates. However, it can be provided, with relative ease, if the solution is in the cloud.

Convergence: Does It Make Sense?
The integration of these two types of data is feasible, despite the issues, but to be honest while I wrote about this topic in the past I never stopped to consider how it would work in practice. Moreover, why would an enterprise want to do it? And what benefits would accrue?

The answer was there in Palantiri's solution, but it didn't manifest itself to me until ThingWorx (a somewhat unusual but memorable name) acquired this company. ThingWorx was working along similar lines, so the spin was that a merger made sense.

Today's software platforms focus on execution processes that generate one of three types of data--unstructured, transactional, or time series. People work in unstructured ways; computer databases such as CRM and ERP employ transactional data; and we can classify M2M data as being based on time and events.

For each of these data types, a specific set of intelligence tools have been developed in order to provide "insight", e.g., a basic Google search for unstructured data. However, the rigid structure of data sets limits the questions that can be answered to those known in advance.


The platform treats all "Things"--people, systems and the physical world--the same way. It enables the creation of a unified environment in which smart things talk to people and each other and where every person or thing can listen and react to every other person or thing.

A New Take on Unified Messaging
ThingWorx’s solution aims to unify data emanating from people with the data that comes from computer systems and the physical world, i.e., M2M devices.

The rationale behind this commendable objective is the challenge of "The Internet of Things", i.e., enabling connectivity to a staggering number of connected devices. Nobody knows the precise figure, but it's set to rise from around 5 Billion to 1 Trillion in the next 10 years.

This challenge has been partially met by decomposing these business applications and by deploying SOA and Web technologies. However, ThingWorx's premise is that retrofitting them to support texting, mobile devices, social networking, and collaboration rarely creates satisfying results.

What's needed is an application platform that collects, tags, and relates the three different data types from these things, thereby creating an operational data store that enables messaging and collaboration between the various users. It allows people, smart systems and smart devices to subscribe to messages from each other.

People can be proactively notified of activities. Smart devices and systems can be made aware of what others are doing.

In addition, a search, query and analysis tool that acts on unstructured, transactional and time-based data simultaneously provides access to this data.

Process-to-Process Communications
Treating "things" the same way allows new processes to be created by connecting up the relevant things in any combination, and they can also be modified as they evolve over time. Modification would reflect usage, e.g., the way they are employed to answer questions and solve problems that had not been anticipated.

In addition, APIs enable integration with other business applications and services, internal or external. And an instant messaging (XMPP)-based remote connectivity model enables secure, real-time interaction between devices, people, and business systems.

The Acid Test: Deployment
This isn't the easiest concept to take on board and it does sound too good to be true. So far there are no named deployments--it's a new development--but a detailed study has been conducted for a provider of automation solutions that help customers optimize their physical building infrastructure.

Right now homes, offices, and commercial spaces are undergoing a smart transformation: connecting and linking environmental sensors, HVAC, office and safety equipment, with external inputs such as the smart grid and weather. This has led to an increased demand for leveraging building intelligence in order to create positive impacts on energy consumption, operating costs, safety, and ROI.

ThingWorx's solution can embrace the diverse inputs that are needed for a smart building ecosystem. For example, bringing the smart grid and weather forecasts into the equation is a recent requirement. But it is needed in order to efficiently query, analyze, and transform building data into meaningful, actionable information.

According to ThingWorx, this cannot be done using traditional software platforms, i.e., those that would attempt to comprehend the relationships and correlation of data from multiple, different sources. Other vendors may disagree, but that doesn't detract from the functionality of this innovative concept.

Summary
Components such as things and people communicate in streams of messages. Other components monitor data, understand it, and send their findings out in messages. This means that applications and people decide which streams to listen to and how to make sense of them.

A connected application is a collection of distributed components that communicate with each other. Some components are the data and functionality surrounding a smart thing or smart device. Others monitor activity and notice if something important happens. They also collect important events and present them to the user. Any one component in a connective application may participate in many different applications.

Connections between components and applications are based on streams of messages so that decisions about which applications are connected to which can be made as new ideas occur.

Bob Emmerson is a freelance writer who lives in The Netherlands. Email: [email protected]. Web: www.electric-words.org.