March 28, 2016
Beyond M2M: Breaking down data silos to enable the Digital Enterprise
First generation M2M systems are running up against limitations when it comes to the Internet of Things (IoT). IoT requires an infrastructure that enables a digital enterprise by supporting ubiquitous, yet secure, data accessibility, system-wide interoperability and composability.
Let’s first address the definition of ‘M2M systems’. A traditional M2M system involves point-to-point (north-south) connections over the Internet between a device and a Cloud service. This is adequate for limited applications like device management and basic data collection and these systems pre-date the advent of the IoT. Unfortunately, these systems create data silos that do not permit peer-to-peer (east-west) data connectivity between devices, or multi-tier system topologies (e.g. distributed computing concepts with edge, gateway, fog or cloud nodes) and thus do not allow the full potential of the IoT to be exploited (i.e. they offer sub-optimal returns-on-investment). Typically these 1st-generation systems only address specific tactical applications and have their own embedded ways of enabling communications – Standards-based or not. As such, they have only generated a relatively small market (compared to enterprise computing markets and the potential of the IoT) and are thus predominantly serviced by relatively small vendors.
Now let’s look at the approach required to fully exploit the potential of the IoT. The infrastructure required is quite different from point-to-point 1st-generation M2M systems. It is being defined by collaborative forums such as the OpenFog Consortium that involve industry giants such as Intel, Cisco, IBM and Microsoft, working with researchers and innovators.
This infrastructure is required to enable a digital enterprise through supporting ubiquitous, yet secure, data accessibility, system-wide interoperability and composability. That is, both north-south and east-west data connectivity to support business value-add through applications at the edge (in devices), in gateways, in fog nodes (appliances) and in the cloud (as remote services).
Many analysts, e.g. IDC, forecast that soon 40% of IoT data will be ‘stored, processed, analyzed at or near the edge’ and that 50% of IoT systems will be ‘network constrained’. Therefore 1st-generation M2M systems (with their dependence on cloud services and always-on Internet connectivity and bandwidth) will not be acceptable for many reasons (latency, determinism, throughput, security, reliability, robustness, recovery, etc., etc.). This has resulted in industry efforts to support concepts like edge intelligence, fog computing and distributed analytics, in addition to, and complementing, cloud services. If 1st-generation M2M systems pioneered the use of the Internet and cloud services to add value to device data, then these 2nd-generation systems are those that will enable the Industrial IoT.
Obviously there are many standards that already exist and are highly relevant to these 2nd-generation systems. The Industrial Internet Consortium has published a Reference Architecture that identifies relevant standards, and the OpenFog Consortium has working groups on open architecture for fog computing using existing standards. Both of these organizations were founded and are led by some of the biggest names in IT, thus with the influence to define the next-generation industrial digital infrastructure.
In summary, if you are looking to implement an infrastructure to support your digital enterprise:
1/ Look beyond traditional (1st-gen) M2M systems. Integrating multiple north-south data silos later will be a nightmare.
2/ Look to industry consortia such as the IIC and the OpenFog Consortium to help you understand concepts such as edge intelligence, distributed analytics and fog computing.
3/ New Standards are rarely required. Applying existing, proven and recommended standards (such as DDS and MQTT) within an open reference-architecture will give you a timely solution with minimal risk.
4/ Consider the reference architectures being collaboratively developed by the biggest names in IT; beware of M2M vendors’ data silos.
Steve Jennis is SVP Corporate Development at PrismTech and recently featured in the Top 10 List of the Most Influential Internet of Things Executives. With over 20 years of experience in high-technology management Steve has achieved success in both large corporate and start-up environments and is widely recognized as a leading business strategist for infrastructure and platform software. Steve blogs regularly on IoT topics and contributes to a number of IoT industry initiatives. He may be reached via Twitter@stevejennis or Email: firstname.lastname@example.org.