The statistics on the number of Machine-to-Machine (M2M) connected devices that will be used in the next decade are astonishing. It has been estimated that this number will top 50 billion, and IBM has projected more than 1 trillion. Open connected computing platforms are helping accelerate the number of devices that can utilize the cloud to communicate and aggregate data. Mostly deployed as edge node and gateway devices to collect and send data, M2M applications will be the catalyst to push connectivity into a myriad of new and existing device platforms. This means that a diverse group of suppliers will find M2M ripe with opportunities to launch new smart services and devices worldwide.

While M2M has been around for more than a decade, streamlining the implementation of M2M applications and services still presents many challenges. OEMs need to be aware of the multiple connectivity protocols and emerging standards they will need to support in any M2M deployment. Additional taxes, tariffs, and customer service issues now come into play along with device security, support, and reliability concerns.

This complexity means that OEMs developing M2M-based systems must look for infrastructure building blocks that enable them to easily engineer platforms that take M2M technology data from the point of collection through the cloud to the point of aggregation. Finding embedded computing resources that simplify this process allows OEMs to provide innovative M2M systems that better serve this growing customer base.

Who is supporting M2M?

Several factors are propelling the explosive growth of M2M applications, and the ecosystem that supports M2M is rapidly expanding. A major benefit of selecting an open architecture embedded computing platform is the rich ecosystem of software partners who have already developed M2M systems. For instance, Intel has developed support software partners for its processors, from Operating Systems (OSs) through middleware. These software companies offer various OSs and tools tailored to M2M market needs, such as horizontal hosting services that can be used for provisioning or updating an M2M system.

Among the group of suppliers who have seen the potential for connected computing are wireless operators, mobile virtual network operators, and Internet service providers who have developed new billing programs for M2M wireless broadband transaction services. Adding to the ecosystem are independent software vendors, also referred to as M2M Application Platform (MAP) providers, who are launching middleware and cloud services for client devices. These MAP services enable secure real-time computing data compilation that supports not only direct communications, but also seamless integration into existing enterprise systems. Besides providing the ability to integrate real-time data, many middleware and framework companies also offer transaction validation and verification services to ensure that billing from broadband providers is consistent with usage levels.

Realistic M2M metrics and usage models

Network operators have historically used the Average Revenue Per User (ARPU) metric to determine a successful service implementation. The ARPU goal has typically ranged from $50 to $60 per user in the United States. This metric is not suitable for M2M deployments. A more realistic goal based on the types of M2M services operators are likely to provide is estimated to be between $5 and $10 per device. This might seem lower than required, but operators can expect the number of customers to grow exponentially due to the spectrum of new technologies that will be needed in many countries, as well as the multitude of devices that will be introduced to maintain the quality of service expected by customers.

A wide variety of usage models is spurring an increasing range of valuable M2M applications that has attracted many suppliers to enter the market. These usage models include:

• New transaction-based billing models from operators, which can facilitate a viable M2M economic environment for customers. The cost of 3G modems is also decreasing, further enabling cost-effective deployments.
• Networks have continued to expand to handle the huge number of low-transaction devices.
• Cost-effective standards-based M2M platforms and general-purpose infrastructure components allow OEMs to simplify system development and deployment.
• M2M software and service frameworks provide autonomous device management.

Trends driving M2M expansion

The trends driving the expansion of M2M devices and smart service development are typically split into three usage model and device functionality categories:

• Indirect nodes and sensors: These provide monitoring services and frequently use ZigBee or Wi-Fi wireless communication, typically communicating via a gateway/aggregator to the cloud middleware.
• Direct nodes: These operate in a stand-alone capacity to send and receive data from the cloud and sometimes use multiple radio technologies to accomplish their mission.
• Gateways: These aggregate data from sensors and deliver commands back to sensors using multiple radio technologies.

M2M devices are being deployed today through innovative remote management systems from horizontal providers who enable over-the-air provisioning and software updates for devices, as well as implement the policy and security to protect data and privacy. In addition, new automatic dashboard services give operational entities access to real-time data, enabling them to make rapid decisions and take actions to control costs and optimize resources. This usage model differs from previous models where dashboards were created manually using reports generated from corporate databases. M2M devices are also becoming integrated into the enterprise operation, eliminating stand-alone M2M islands that do not adequately support corporate database integration.

M2M networks provide an advantage in that collected data becomes actionable data that can be used in real time. Going well beyond the initial M2M smart meter deployments in the energy industry, M2M devices are now transforming a broad range of infotainment, automotive telemetry, and gaming deployments, as well as applications in the industrial and building/home automation markets. The health care market also has the potential to benefit from M2M, as the technology could help spawn subset markets such as in-home patient care versus in-hospital care or assisted-living options.

Resources to streamline development

Production-ready systems that are based on open standards-based platforms are the building blocks that will help OEMs simplify smart service deployment opportunities. This type of proven approach saves OEMs from the complexities of mastering the breadth of communications technologies and standards, allowing them to reduce development time and design risk. It also allows an OEM to test a smart service application in a connected environment that will be similar to an actual deployment.

To show how smart service developers can use this type of resource, consider the Kontron KM2M806 (Figure 1) as an example of a deployable, service-ready system. Featuring application-ready middleware, the hardware platform supports multiple types of connectivity out of the box for development purposes and testing, including Ethernet for wired broadband access, 802.15.4 for wireless personal area networking, Wi-Fi for wireless local area networking, and 3G for wireless broadband networking.

Successful M2M deployments also need additional software and services to ensure that data gathered by the M2M nodes can be processed and actions can be taken. Middleware providers have made this easier for OEMs by offering suites of MAPs that integrate the real-time data collected into business databases to minimize human error and automate audit trails. MAP companies provide middleware and service Software Development Kits (SDKs) that can be used for both development and deployment.

Development SDKs are used by companies who want to develop their own code and integrate a complete solution. Device APIs and cloud application APIs are provided to develop agent code and application connectivity. Toolkits are provided to integrate and troubleshoot connectivity with cellular modems. In contrast, deployment SDKs provide a complete turnkey solution, which typically includes cellular connectivity and existing relationships with carriers. Coding is not required; only configuration is needed using a graphical design tool that speeds deployment of a field trial prototype.

As the M2M marketplace expands, it is important for OEMs and service providers to adopt proven and robust M2M systems and building blocks that are optimized for market use. The ability to simplify the complex M2M supply chain allows developers to spend their valuable resources engineering innovative M2M systems and services, enabling them to get to market quicker and realize faster time to revenue.

Satish Ram is product manager at Kontron.

Kontron www.kontron.com

Follow: Twitter Facebook Google+ LinkedIn YouTube