M2M: Connect for Smart Services

In the new M2M world, build awareness and connectivity into your product and be ready to reap the rewards of being a smart service provider.

The purpose of the term M2M (machineto- machine) is to cover those forms of data communication not covered by other mainstream consumer and offi ce oriented systems. Traditionally this has meant telemetry applications, where for example a sensor may interact with a server to generate an alarm. However, the introduction of “wireless M2M” has opened up an almost unlimited number of new application possibilities, ranging from car telematics through to personal security and even telecardiology – where an implanted pacemaker can be monitored in real time from anywhere in the world.

The term “Pervasive Internet” implies Internet everywhere and the connected environment that results from this. It emphasizes many connected devices acting together in a systematic way, rather than individual devices used on their own. It also implies network-based processing and reporting of data and is less concerned about the means by which devices are connected to the Internet. It also implies systems interacting with oneanother and sharing data, across the Internet.

There is considerable overlap between these terms, although Pervasive Internet has a stronger emphasis on volumes and systems and M2M has a stronger emphasis on the connection between machine and network. To avoid confusion, therefore, Harbor Research has adopted both terms and this article refers to “PI-M2M” as the collective term for connecting machines and sensors through any means to the Internet, processing, reporting and sharing this data as appropriate and automatically monitoring/ controlling any number of these machines remotely as appropriate.

There are now a variety of technology alternatives for connecting machines for PI-M2M applications, sometimes as competing alternatives and sometimes as complementary elements of a complete solution. It follows that a study of just one of these alternatives – for example cellular - will not provide a view of the whole PI-M2M opportunity and that championing just one technology will tend to over-emphasize its opportunities.

To overcome this, Harbor Research takes a broad view of the market, assessing the market opportunity of the devices and machines themselves that are to be connected and then analyzing the alternative technologies for connecting them. This minimizes the potential for double counting of device populations and rationalizing the forecast so that it refl ects the likely take-up rate for all the main technology alternatives.

Going past the product

Any industrial manufacturer that has not woken to the fact that it must become a service business is in serious peril today. And sadly there are such businesses. In fact there are many of them – companies that still think of themselves as the builders of things and that state their gross margins, operating profi ts, and other measures of success solely in terms of “the product”.

But even their more enlightened competitors, the ones that have begun to wrap valuable services around their products, and in some cases profi t directly from those services, are enjoying only a temporary advantage. They may be improving their customer relationships by taking on various maintenance and other burdens, but that will only take them so far. And that is because a select group of companies is already upping the ante again. Soon, it will not be enough for a company to offer services. It will have to provide smart services.

Smart services go beyond the kinds of upkeep and upgrades a manufacturer may be bundling with its product today, both in its value to customers and in its cost-effi ciency to the manufacturer. But to be in a position to provide these, the manufacturer must build intelligence – that is, awareness and connectivity – into the product itself. And it must be prepared to act on what the product then reveals about its use.

As an example, consider Heidelberg Druck maschinen, the famed German maker of high-end printing presses. Throughout its history, the company has offered repair services to customers. When it developed the ability to monitor the equipment remotely, it could provide maintenance much more cost-effectively. Today, with its machines communicating continuously over the Internet, Heidelberg has the access and insight to optimize their performance in the customer’s shop.

The total product life cycle support Heidelberg now offers – extending even to deinstallation and resale – represents a step-function change in value. It is the network context that makes the difference and creates true intimacy with the customer.

The rewards of becoming a smart service provider are hard to deny – Harbor has documented organic growth rates in double digits for many companies that follow this path. The leaders are establishing the new performance benchmarks for their industries, deriving more than 50% of their revenue, and 60% of their margin contribution, from services as opposed to product sales. For most management teams in product-centric companies, numbers like these sound like nirvana.

Adopt a new perspective

Joining the ranks of smart service companies is not primarily a technical challenge. The technology, while critical to the task, is well enough established by this point. Rather, in most companies, the biggest challenge will be in getting management to adopt a new perspective on the nature of the business. The companies in the vanguard of smart services think differently about what they exist to do and how they profi t thereby, but theyhave come to that new heading by degrees.

Smart services are a wholly different animal than the service offerings of the past, and customers perceive them as having entirely new value. To begin with, they are fundamentally pre-emptive rather than reactive or even proactive. Pre-emptive means that actions are based upon hard fi eld intelligence; you launch a preemptive strike to head off an undesirable event when you have real-world evidence that the event is in the offi ng. Smart services are thus based upon actual evidence that a machine is about to fail, or that a customer’s supply of consumables is about to be deleted, or that a shipmentof materials has been delayed, and so on.

For customers, smart services create a new kind of value – the value of removing unpleasant surprises from their lives. Meanwhile, because the fi eld intelligence makes product performance and customer behavior visible as never before, the supplier gains unprecedented R&D feedback and insight into customers’ needs, and isthus able to provide even greater ongoing value.

Finally, because it is impractical to deploy human beings to gather and analyze the real-time fi eld intelligence required, smart services depend on“machine intelligence”. In a smart services environment,reliable and fast microprocessors do what they areparticularly good at doing: digesting large numbersof data points, talking to each other about the data,controlling each other based upon the state of the data– all in a matter of nanoseconds.

Human beings cannot do this, nor should they; this incessant stream of ongoing business information should be invisible to people. At the same time, all this invisible machine activity makes the state of (i.e., the information about) a business’s assets, costs, and liabilities vastly more visible to managers and to the decision-makingprocess – when decision-makers need or want to know.

This is not dazzling future speak; for many companies it is already reality. For many more, it is a matter of reaping the harvest of seeds already sown. For decades, we have been steadily building electronic intelligence into manufactured objects by means ofsensors, controllers, and microprocessors.

Today, virtually all products that use electricity– from toys and coffee makers to cars and medicaldiagnostic equipment – possess inherent data-processingcapability. And each has a wealth of information tooffer about its status, usage, and performance. So if amachine, or a consumer product, or a building is notpresently monitoring every detail that its creator mightwish to extract, it can easily and cheaply be made todo so.

Business benefits

 

The benefits of smart services vary for different companies, and for different industries. But in general, the potential benefi ts for a particular company are anyor all of the following:

• Cost & service time reduction
A product, such as an elevator, breaking down canimmediately advise its service center (which could bethe manufacturer) of this as an alarm. The servicecentre can then interrogate the product remotely tofi nd out what happened and ensure that techniciansare equipped with the correct spare parts beforemaking a visit to site, thereby reducing the numberof site visits and technician time whilst improving theservice response time.

Vending machines can signal that they are short of stock and need replenishing. Ticket machines can indicate when they are full of change and need emptying. These actions obviate the need for constant checking while at the same time ensuring maximum availability of the service the machines are offering.

• New or better services
With performance data available at any time for their products working on site, service providers are in a better position to predict possible future failures (e.g. bearings need replacing) or to advise on a course of action to save cost (e.g. interval metering showing peak usage trends).

And combining GPS with wireless enables tracking of people or assets. Insurance companies are experimenting with selling individuals car insurance with a deep discount if customers allow the company to monitor their driving via GPS and a wireless link. In many cases, individuals seem to be willing to trade privacy for a lower rate.

• Speed & flexibility
A wireless billboard system can introduce changes instantly to many different billboards at once, offering a service timed for peak viewing of the target audience. Vending and ticketing machines can be updated with new software, pricing and displayinformation at any time at low cost.

• Building relations
Customers typically buy products and are then on their own. Instead, the manufacturer or supplier can maintain an ongoing relationship and create input for a CRM system, for example. This might increase the chances of a car owner buying the same make next timearound, for example.

• Additional sales
A connected photocopier can sell paper and toner. A connected vending machine can cater for credit card purchases, opening the prospect of higher value sales such as CDs and videos. In the toy market, products could sell upgrades for themselves. For example, dolls whose personality matures over time.

• Stronger product offering
Rather than selling just a product, companies can sell a solution or capability. Cars that self-diagnose and warn of potential diffi culties before a long journey is undertaken; washing machines that can predict their own breakdown. These are examples of sensors connected wirelessly to larger systems that can then process the sensor data and provide a better support service.

• Advertising space
A connected product can be used as an advertising medium. Vending machines equipped with displays can show advertising, local information of interest, special deals or even fi lm trailers, perhaps immediately after a purchase is made. A particular advantage of this channel over traditional media is the opportunity to target a message according to the target group, time of day, etc.

• Collecting valuable information
A machine can be used to collect information that might have a commercial value. A product development function can gain immediate feedback on how a particular product is used, which features are most popular and what problems typically arise. This can shorten the lead time to an improved or updatedversion, providing a competitive edge.

Market drivers & forces

For every piece of equipment, there is a company that both needs to monitor or maintain it and occasionally needs to adjust its settings. Companies spend a great deal to monitor equipment manually and, more importantly, lose a great deal in aggregate to alarms, malfunctionsor leaks.

This story is repeated in many industries. It is becoming less sensible to send technicians or meter readers just to retrieve simple pieces of data. The high cost of transportation means even small gains in effi ciency can translate into signifi cant profi t improvements, and staff can be freed up for higher-value functions.

When it comes to service, customers now expect better and in less time. For companies, this implies demand-based, just-in-time servicing, which translates to running systems closer to capacity, managing resources effectively and improving productivity while cutting costs. Using PI-M2M applications to make sure the data is in the right place at the right time is an important step in this direction.

Networks, whether unifi ed or virtual, are providing greater one-stop shopping and further improving operational efficiencies. End-to-end systems, which were once available only through private systems, can now be tailored to specifi c industry requirements.

Then there is the area of policy, regulation and legislation. Compliance concerns particularly affect utilities and have been increasing the demand for AMR (automatic meter reading) in electricity, water and gas markets. Regulators in the utilities sector are often concerned with leakage, safety levels and billing accuracy.

It is also becoming less economical to devote personnel to simple but essential tasks such as meter reading or checking on machinery.

Legislation, such as requirements on the monitoring and recording of food temperature or environmental emissions, puts a legal requirement on some industries to record and produce an audit trail for track/trace purposes. Recent incidences of BSE have raised the need for tracking batches of a wide variety of foodstuffs and other products, such as toxic chemicals. Automated methods improve safety, eliminate the need for humanintervention, reduce costs and improve reliability.

Trends & implications

Based on all the knowledge gathered about supplier progress, technology, and customer adoption, there are several major trends which have wide-reaching implications. Harbor Research strongly believes these forces will deeply affect business strategy and customer relationships in the near term. Some examples of forcesdriving this change:

• All electronic devices will eventually be connected to intelligent networks and the Internet. Technology improvements will make it cheaper and easier to implement connectivity solutions in everyday devices (beyond servers and PCs into machines, toys, etc.). Devices will act as gateways or beachheads to consumers and businesses.

• Suppliers will want to be as close to their customers as possible, and users will place greater emphasis oneffective service and support.

• Electronic devices will have rapidly increasing intelligence and functionality.

• Demand for adaptability, agility and features will grow. Sensors, in particular, will grow signifi cantly in types, applications and capabilities.

• Communications options will increase, have greater bandwidth, and be cheaper (free Internet access). Phone lines, Data lines, Power lines, Cable and Fiber Lines, Cellular, Satellite, Radio Frequency options will abound, all in service to providing communications and connectivity.

• Electronic devices and systems will be self-adapting. As witnessed by computers beating chess pros and with automatically adjusting car seats, the digital nervous system will increasingly be of service to users, not the other way around. Companies will exist to house, facilitate, and recommend better options for dealing with individual needs and tastes, where people will opt for service contracts instead of buying products outright.

• Customers will creatively apply and integrate technology in their work and personal lives to unimaginable levels. The technology will become increasingly become invisible to the user.

• Customers will require a “tailored experience” from service providers. Not only do customers expect suppliers to anticipate their needs, they will wantsuppliers to project an experience for them.

Developing a strategy

The World Wide Web is both a helpful and a very deceptive proxy for the new world represented by the Pervasive Internet. When the Web came along, the only thing known for sure about it was that its potential was huge and the business stakes high. Because adoption of the Pervasive Internet has been somewhat slower, perhaps companies understand a bit more than this as they move forward, but there is still large potential for gain and loss, and many factors that are not well understood. Moving early is very important – it’s possible to achieve real dominance by moving early, and to be left behind bydelaying.

Many of the same marketing methods are also valid. The Web gave us viral marketing, and this will still be a powerful marketing concept within the M2M universe. And because some M2M applications will have human interfaces, and many of these will be Web-based, Web marketing techniques are highly valid, at least for these applications.

Partnerships and portals are very important. Personnel will be specialized in fi nding and evaluating potential partners, and cultivating them into actual ecosystem participants. Portals (which always included search functions) were the fi rst killer apps of the Web. They will not be fi rst in M2M, but they will be important. And products will be applications. Harbor refers to network-enabled products and services as applications. Indeed this is what they are, in exactly the way IT has always used the term.

For most players, PI-M2M planning will be like much high-tech strategy, in which down-in-the-weeds technology considerations will have to be made sideby- side with the broadest corporate “vision things.”

Finally, just as PI-M2M will rock companies, it will rock individual careers. Many individuals will be unable to adjust to the required new ways of thinking, and very few will be able to get their mental arms around the full complexity of it. Thus an important consideration in strategizing is who to invite to the table. A breadth of points of view will be necessary to achieve a complete vision, and it will be a serious challenge not to let the diversity of viewpoints result in paralysis by analysis, or a protocol of excessivelypolite listening.

Harbor Research Inc. (www.harborresearch.com) provides strategic consulting and research services in PI-M2M – the convergence ofembedded device computing and global networking.

Designed for Wireless High performance embedded processing power plus integrated wireless connectivity provides a boost for M2M application developers. Providing industrial product developers with powe r f u l central application processing and built-in support for global cellular (GSM/GPRS) wireless internet connectivity, backed by a comprehensive operating system and application software development environment. That is the aim of the Wireless Microprocessor from Wavecom, a new breed of embedded wireless computing cores. Along with high performance embedded processing power with wireless connectivity in a single surface mount component, in combination with Wavecom’s Open AT software suite, the Wireless Microprocessor claims to provide the lowest total cost of ownership for any industrial wireless machine through reducing the overall number of components and shortening time-to-market. Using Open AT software, developers create machine-to-machine applications in C-language that can be maintained remotely, over-the-air, throughout their lifetime. Open AT OS also allows developers to execute their program natively, without the need to interpret code, requiring a minimum memory footprint and processor resource overhead. Case Study: Automatic Metering In the case of an M2M metering application, the meter needs to constantly measure the energy quantity used, store the related consumption data, and then transmit the data to a back-end server at regular intervals. The data is used to manage peak consumption, calculate accurate bills and/or perform preventive maintenance. Consumption information can be sent either via SMS or GPRS. Additional information can also be communicated as alerts in the case of prepaid meters when the level of authorized consumption is reached. In conventional configuration, the meter has an external processor connected to several peripherals to drive power measurement circuitry which converts the energy signal giving consumption data and stores it in an external memory such as EEPROM. It also controls interface devices like keyboards and displays. For GSM communication, the external processor communicates with a Wireless CPU (using AT commands), which then sends the SMS or opens a GPRS connection. However, the Open AT Software Suite running on the Wavecom Wireless Microprocessor eliminates the need to use an external processor connected to several peripherals to drive power measurement circuitry, nor to monitor the peripherals or manage any unsolicited event. The consumption data is stored directly in the data storage area of the Wireless Microprocessor, which also directly controls input, display devices and external connectors in addition to the wireless GSM communication. The alarm signal is directly connected via GPIO (General Purpose Input/Output) of the Wireless Microprocessor, and an embedded Open AT application monitors the signal and sends an SMS in case of alarm detection. Reducing TCO This application shows how the functionality of hardware components can be optimized by using the Open AT Software Suite, effectively reducing the total cost of ownership (TCO) through four key elements: 1. Material cost: removal of host CPU, analogue & digital ASICs, display drivers, discrete components; fewer parts i.e. connectors/cables, PCBs qty, and overall size reduction. 2. Time to market: saving lost revenue incurred by extended product & custom tools development. 3. Design costs: via use of Application notes, reference designs and a range of Integrated Development Environments (IDE) such as Eclipse. 4. Post-deployment maintenance: investment protection via DOTA & system recovery. Wavecom, an active player in the M2M sector, says it is committed to continuously introducing new ideas, features and products to allow customers to innovate in accordance with market demand. Based on information from Wavecom (www.wavecom.com).

Getting Real With M2M A very promising concept but not yet true reality, says Amit Mehta, who details his experience of an actual M2M project. Connecting remote physical assets to the enterprise has existed for many years under different terminologies. These have assumed new forms with the emergence of three major trends: bandwidths going everywhere; open standards prevailing; and computing power doubling/tripling every year. Today’s buzzword is M2M (machine-to-machine), signifying an era in which machines talk to machines to guide activities. This is a contrast from people talking to people, the emphasis of the PC era. Although M2M is today being piloted in some form or other across industries, we still not seeing mass adoption and deployment. Besides business challenges such as low level of management understanding, technology challenges still remain. Take the following case. Moblize Advisory Services, which specializes in helping clients exploit the real benefits and select, build and deploy best-of- breed M2M technologies, was engaged by a Fortune 100 energy company looking to garner benefits of M2M in action, e.g. reduced downtime, asset tracking, enhanced productivity. Assets to be monitored included tanks, pipelines, wellheads, motors and gearboxes. Reviewing the client business processes, it was established that in addition to lowering costs, the client could potentially launch innovative business models to capture new revenue streams on the growth side. Although project implementation was successfully completed and the next phase is being discussed with the client, as the project progressed in phases, we encountered challenges which started to expose the realities of M2M in action, as summarized below. Data collection technologies This layer today encompasses many technologies (Zigbee/ Bluetooth/UWB/RFID etc.) that are competing for market share. For this particular project the focus was on wireless mesh sensors. We discovered during the project that many manufacturers claim in their marketing literature to be “Zigbee ready”. This really means that only the chipset for the RF layer in Zigbee complaint, but no other parts of the Zigbee protocol are yet in place. Many manufacturers don’t support industrial applications– e.g. “wireless” RS-232 and USB “cables” for “office use only”. A few did have some “sensors” for industrial usage but the UL certifications and packaging are far less than“industrial” grade. Data travel technologies This encompasses all wireless LAN and wide area network technologies, which come with its own set of challenges. During the project, the focus was selection between Wi-Fi /Cell modems/Long range mesh radios. Wi-Fi, though mature and secure now, still gives chills to corporate security policy makers; cell modems are famous for random drop in packets; mesh LAN are more focused on campus applications than longer range. Data destination technologies This encompasses application software (from a variety of vendors) traditionally built as proprietary technology using client server architectures. The new generation of technologies are Web based/support Web services, but they must coexist with the older generation technologies. Integrating the two involves writing custom interfaces. You could keep as two different information silos but that would defeat the idea of seamless access to anytime/anywhere information, not to mention high costs. User presentation technologies Encompassing platforms like PDA, cell phones, laptops, tablets, desktops, based on variety of OS like Symbian, CE, Mobile 2005, J2ME, challenges faced in this arena involved lack of drivers to support viewing the data on different devices. In our case, the hardware manufacturer selected decided to stop manufacturing that product line and migration to new platform became an IT headache as the handheld application had to be recompiled. Common challenges Challenges experienced across the board in all the four layers included lack of one-stop shop specialists to install/ troubleshoot the equipment in all four layers; power-hungry equipment like mesh sensors and IP satellites discouraging the use of batteries or solar panel for more frequent use; wireless interference; proprietary nature of devices. In our opinion at Moblize, although subsets of M2M, like SCADA, AMR, are possible, large scale deployment where anything can talk to anything is far from becoming a true reality. Besides the obvious technology challenges, which over time will be addressed, business challenges remain. For example, real decision makers don’t go to the Zigbee, Lon, or Bacnet standards meetings. And in large corporations, where most of the potential early adopters are, procedures/ policies and myriad bureaucratic challenges can dramatically slow down adoption. For users, little material is available on aspects such as business process transition mapping to leverage on M2M technologies. Clients are asking about how to make money through connecting their remote assets to the enterprise. Consulting in this space is an opportunity – helping clients think of benefits around revenue generation versus pure cost reduction (taking 50,000 ft view and then executing at 50 ft), and this is a growing business for Moblize. Amit Mehta is Principal of Moblize (www.moblize.com), which specializes in helping clients transform their business and processes by connecting remote (fixed/mobile)assets to enterprise. For more insights: http://machine2enterprise.blogspot.com.