Smart machines and connected objects | March 24, 2009

The next major disruption for the Internet may come from the potential connection of billions of machines (with M2M) and trillions of objects (with RFID and the Internet of things). It assesses the key technologies to accelerate such a development, the applications that are or will be the most relevant in a connected mode and the early positioning of the industry players on this promising market.

M2M and the Internet of Things

Compared to Machine-to-Machine (M2M) technology, which links up machines having their own intrinsic value and uses outside communication, the ‘Internet of Things’ goes an important step further.

In network mode, it analyses labels placed on specific objects and, knowing the objects’ location and status, provides the basis for new services.

Whilst its functionalities are similar to those of M2M, the new paradigm of the Internet of Things is more complex and will require the roll-out of a new infrastructure.

In both cases, the market will develop in large-scale vertical segments, once the proper sector organisations have been put in place.

There are countless opportunities in the markets of both M2M and the Internet of Things – initially in terms of cost reduction, and then through the emergence of new services.

Internet is gaining a new life with billions of machines connected

The Machine-to-Machine (M2M) revolution has meant that machines in communication with each other become just a bit more intelligent. Humans are no longer alone in being able to connect to the Internet: now any kind of machine can do so.

• It is a sizeable market indeed, with more than 2 billion such machines in the world, 630 million of them in Europe. Thus far it has developed primarily in the fields of internal, wired applications, but the most dynamic surge will come from wireless M2M where cellular, satellite and other solutions will open up new, otherwise impossible opportunities for roaming modes and in open environments.

• In 2007, the cellular M2M market was worth EUR 6 billion, comprising more than 11 million active modules. Its rapid growth rate will sweep it to 61 million modules by 2012, worth some EUR 38.5 billion.

• There are many winners among a wide range of technology players, with M2M being not a technology in its own right but a compilation of solutions from the software, electronics and telecom sectors.

Financial appeal underpins M2M growth in some sectors

The initial attraction of M2M has been in cost reduction and optimisation. With distance monitoring, a machine user can save on costs of travel, bad data recording and machine downtime. Several industrial sectors have used M2M to cut costs in adapting to regulatory compliance.

• With distance data reading, there are gains over and above cost savings. M2M can lead to innovations in customer service throughout machine lifespans and/or new pricing models based on real-time use, taking into account weather conditions or distances travelled. Such gains, obviously, require that solution providers and machine manufacturers change their business models and adopt new technological solutions.

• On certain machines and applications, M2M is already well-developed. The most advanced sectors are in transport, with fleet management and such telematic systems as safety, theft prevention and dynamic navigation, and in utilities, principally for meter readings. Other widespread applications are found in security services, with alarms to protect property and buildings, and in electronic money services – benefiting in particular from the shift to cellular communication.

• Amongst other promising segments of the market are health, advertising (with electronic billboards, and point-of-sale signage), ticket dispensers and even toys and gadgets. As with most other solutions which target the public at large, the business models for these new sectors have yet to prove themselves, given the frequent reluctance of the end-user to pay.

Moving into all sectors, beyond industry and into the invisible

Its inherent technological complexity for long thwarted the development of the M2M market. At the outset, it was limited to proprietary applications for internal use in a handful of specific sectors where the financial gains were obvious.

• Seeing how different sectors were separately opting for very similar functionalities, various ICT players and start-ups set to developing new common solutions, primarily based on mature technologies (sensors, GPRS, XML). They could thus avoid issues of complexity and benefit from lower (amortised) technology costs. Some even developed entire platforms.

• The perspective of M2M users – machine makers and solution providers – is, however, less technology-driven being vertically focused on sectors. Sales are organised sector by sector, with solutions being adapted and pre-packaged for each one, even though the same application is used in all of them.

• The general trend, then, has been for progress to be made first in vertical segments on internal projects aimed at cost savings. Later, developments have spread to other, more complex areas of activity in open environments, and even to the general public, initially with the end-user’s machine being operated by its manufacturer or service provider and ultimately being run directly by the user. These public solutions are struggling to get off the ground, but they could well do so if successful business applications of M2M can be adapted to residential use with, for example, meters and alarms, and thus open the way to home automation, also known as domotics.

Still facing serious financial and management challenges

Enhanced performance levels, attractive for new clients, could come from several technical improvements in safety, reliability, latency, network coverage, working life and more. The real stakes, though, are not in the realm of technology.

• Cost is key, as it is on many a market. The prime issue is the level of upstream investment in software and modules. These fixed costs can only be earned back on large-scale projects, which require a strong commitment on the part of the client, or user, to M2M.

• The M2M is particularly fragmented, divided between numerous technologies and players, often niche ones, operating on various links of the value chain. Some players market proprietary solutions – given the high costs involved, this inevitably slows down any rapid expansion. The M2M ecosystem is populated by a large number of new players, and comprises several hundred providers of complete, or partial, systems. From the point of view of the end-user, a good dose of streamlining is essential.

A few key players are shaping the ecosystem

Now, two, complementary, dynamics are cutting down the clutter of the fragmented ecosystem. Several major players, Siemens and Orange among them, are each building alliances, making for easier interaction between the various entities involved. The fruits of ongoing consolidation in the industry, in particular over the last two years as witnessed by Wavecom/Sony-Ericsson and by Qualcomm/Nphase, are also shrinking this fragmentary element.

• Consolidation is at play especially in the electronics industry, which is faced with tumbling prices for modules. Some players are adapting to this process of commoditisation with low-cost packages. The larger ones, rather, are seeking to resist it by adding on advanced software, both in modules and in middleware, in terms of data management on remote servers. In an increasingly competitive context, their offerings, ever more integrated, have started off with integrated circuits capable of directly incorporating the SIM card. Siemens (now part of T-Mobile) and Wavecom are the dominant players, with Enfora and Telit their chief challengers.

• Mobile and satellite operators are having to deal with market conditions far removed from their traditional battlegrounds – among them poor use levels of unitary networks and low ARPU. Nonetheless, M2M does represent real opportunities, given the low churn rates on the markets and the large scale of contracts, when a single client can mean sales of thousands of modules. It can also lead to new growth points, with 4% of the revenues of mobile operators in 2012 coming from Europe, and 12% of their user population worldwide. Their traditional markets are saturated, and they are tending to use more surplus capacity on such infrastructures as GPRS, and satellite.

• Among the mobile operators, very different positions are being adopted. Most seem content to provide basic connectivity packages with rate plans, operating in their traditional business sector. Certain others, such as T-Mobile UK, contract this out entirely, allowing specialist MVNOs such as Wyless and Jasper Wireless to enter M2M or to operate in a vertical segment. There, they seek to stand out by touting the broad geographical coverage which results from aggregation or by such advanced services as static IP. A small number of more sophisticated players have grown vertical, integrated packages, notably SFR in electronic money transactions, BT in security, and Telenor in electric sensors.

A radical upscale for Internet with connected objetcs

The moment is fast approaching when it will be possible to get any object to communicate, including ones without electronics – thanks to recent technological developments.

• A whole raft of innovations in electronics will serve as catalysts. Circuit miniaturisation will allow contact with objects of any size, and advances being made with batteries will guarantee them a long life without recharging. Alternative solutions such as e-paper will allow objects to function as billboards. And a wide range of sensors can collect a whole swathe of data, including physical and geographical.

• The most unassailable innovation of all is RFID. Using HF and UHF technologies, passive RFID tags can remotely provide a consultable identifier without using any energy. These are the clearest type of identifier technologies for the user, enabling the ‘Internet of Things’ to emerge. Other forms of labelling can be envisaged, at least in a transition phase, in the form of classical bar codes, 2D barcodes such as QR, and serial numbers, such as the ISBN, which can be analysed photographically. All these are less attractive solutions than RFID, since they require short distances, must be in line-of-sight and can only address an object by its category, and not by its specific individual value. Further, the RFID can be used on uneven surfaces, and, in its guise as a chipless tag, be easily affixed to, or printed on, any object.

New applications from the Internet of Things

The Internet of Things will expand the classical Web by locating an object and its status through a mobile terminal reading its codes from labels, ideally using RFID.

• New applications could emerge from the ability to complement and compare information, by linking the code in an object’s ID tag with a database of codes and with various programs, through an Internet connection. In the food sector, for example, it would be possible to automatically collect contraindications on a specific product, in terms of allergies, dietary factors or the presence of GMOs. The tourist sector could benefit from visitors accessing extra information about a location. And a flagship application could, no doubt, be in comparing prices for consumers, replicating the success of current Web-based price comparison services.

• The underlying logic of all these potential new uses is more one of improved customer service on physical products than one of defining new and independent services.

A new infrastructure is needed

Unlike M2M’s use of existing infrastructures and mature technologies, the Internet of Things is constructed with new technological building blocks at all levels. A handful of existing resources are also used, such as the way the Web holds some information on products and people, although not on product identifiers.

• Identification is the first, essential step, with all objects being fitted with labels – ideally RFID. Even this technology is not fully mature in certain frequency bandwidths, and can lose quality in the proximity of water and metal and from simultaneous readings. As yet, too few objects have been fitted, and then mainly in internal projects.

• Equally essential is the development of data standards, for allowing uses in open loops, with one object possibly being used in any number of situations. In this regard, some excellent progress has been made through EPC initiatives, even though it is still lucrative for some players to use other codes.

• It will be necessary to allow the exchange of data in open environments, by creating a ‘parallel Web’ of objects. The EPC has developed the EPCglobal Network, with its PML language and associated Object Name Service (ONS) addressing services (the equivalent of DNS with IP). It features Discovery services (equivalent to search engines), Information services (read: databases) and security facilities (equivalent to login/password and rights management systems). Some industry players have started to position themselves around this approach, including Verisign and Orange for ONS.

But first: a sound business model

The required infrastructure is not, even technically, in place. The real problems seem to be more financial and management-related than technical.

• For products such as foodstuffs and medicines, the unit costs of RFID tags are too high, in relation to their value. On top of the cost of the new infrastructure and RFID (borne solely by the manufacturer), there are many other costs which will, at best, be met only in part by the consumer – these include the reader in the mobile phone, applications and EPC addressing.

• The driving force in the business model of the Internet of Things is more one of the service product model (servuction) than product sales. Before the market can develop properly, some industry players will have to redefine their own business models. Clearly, the major cost savings are at the end of the logistics chain, for distributors. The actual producers of various objects are not rushing to take up RFID, despite the exhortations of top-end distributors, with applications limited to pallets and not touching specific objects. To do so would principally raise costs, with little new revenue. Nor do producers want to label their products, not when the most likely beneficiaries of these new services will be third parties using the application to promote such services as price comparison – at the cost of producers.

• The Internet of Things is no doubt one of considerable promise, but it will probably take many years before it expands beyond industrial applications. It differs fundamentally from M2M, which is best seen as a transition, in that the shift to the new paradigm of the Internet of Things will need much investment at all levels without any clarity on the returns on this same investment. In fact, it is probably only when retail or industrial branch organisations, and other such players as tourism bodies and environment agencies, get themselves involved that the Internet of Things will have reached the point of takeoff.

Vincent Bonneau is Project Manager, Machine-to-Machine, for IDATE. Michael Nique, and Samuel Ropert, Senior Consultants, IDATE, also contributed. For more information, please visit www.idate.org.