Changing the Way We Work

Cisco Virtualization Journal

Subscribe to Cisco Virtualization Journal: eMailAlertsEmail Alerts newslettersWeekly Newsletters
Get Cisco Virtualization Journal: homepageHomepage mobileMobile rssRSS facebookFacebook twitterTwitter linkedinLinkedIn


Cisco Virtualization Journal Authors: Lana Tannir, Liz McMillan, Evelyn de Souza, Elizabeth White, Carmen Gonzalez

Related Topics: Cloud Computing, Cisco Virtualization Journal, Big Data on Ulitzer, Internet of Things Journal

Blog Post

Fog Computing and the 'Internet of Things' Analytics Hardware | @ThingsExpo [#IoT]

Fog Computing and the Growth of IoT Analytics Hardware in 2015

Fog Computing and the Growth of IoT Analytics Hardware in 2015

Internet Of Things - Hardware Play a Major Role: We see a great increase in support from software platform vendors for IoT implementation in enterprises, especially from bigger players. Up to now most of these support comes in the form  of  software frameworks, platform, libraries and components, which range from :

  • Agents : Help to connect different devices and source data from them.
  • Hubs : Provide the ability to queue messages from multiple agents and ensure that the agents are not waiting for the consumption
  • Stream analytics: Provide real time processing of streaming data that come from hubs.
  • Big Data Engines : The MPP platform that help to persist the data and process them
  • Machine Learning Engines : The software model that provide machine learning and predictive analytical capabilities
  • Visualization: The layer which provides meaningful insights of the machine learning data.

As evident all the above software components are important in the life cycle of  IoT. However the hardware components do play a major role in IoT enablement  due to the following things.

  • IoT is all about speed and real time. Most IoT use cases are not of much use if they the decision happens at a later time. Hardware components play an important role in processing life cycle in terms of speed and performance.
  • IoT decision making process is as local as it is centralized. This means that IoT processing has to depend on decision making in two distinct places.
    • One in a centralized Cloud repository where the over all the insights from across the devices from a larger industry perspective has to be taken. This is some thing like a Over all decision making in a Product Life Cycle Management managed by multiple devices in Shop floor, assembly and inventory points.

o   More important a localized decision making process that predicts the events  that quite  local to its place and takes a quicker decision much before a centralized cloud repository could realize them. One example could be a monitoring  device tha  monitors the temperature levels of  a plant machinery and takes decision about  cooling measures, this kind of decisions cannot wait till the centralized cloud process finds them, as by the time the plant could have stopped  functioning due to equipment mal function.

  • Unlike Big Data processing which is one way , i.e. from source ‘on premise' location towards cloud, IoT processing is Bi-directional which means the origin of the source of data has to constantly receive the information back from decision makers and act on it. Due to the increased security fears associated with IoT in todays world, most organizations won't be comfortable with a reverse flow of decision from Cloud back to the devices directly, rather would be comfortable with an intermediary hardware that is fully controllable at the source of data and where the decisions from the cloud server are handled and processed.

All these points indicate the importance of Analytics Hardware as part of IoT processing life cycle.

IoT  Analytics Hardware: The concept of having hardware components in the life cycle of major processing scenarios is not  few,  in the past the following hardware/appliances were used to augment the lifecycle of software application  process.

  • XML Appliances As Part of SOA Life Cycle: An XML appliance is a special purpose network device used to secure , manage and mediate XML traffic. IBM WebSphere DataPower XML appliance is one such implementation of device, where it support security , pre processing and acceleration of XML messages.
  • Integration Appliance As Part of Hybrid Cloud: Some vendors pushed for a local integration appliance while integrating On Premise applications with their Cloud equivalent. For example , IBM® WebSphere® DataPower® Cast Iron® Appliance XH40 is a self-contained, physical appliance that provides what is needed to connect cloud and on-premise applications.
  • Security Offloading & Validation As Part of Load Balancers: SSL offloading relieves a Web server of the processing burden of encrypting and/or decrypting traffic sent viaSSL, the security protocol that is implemented in every Web browser. BIG-IP®Local Traffic Manager with the SSL Acceleration Feature Module performs SSL offloading.

The above examples clearly point to the use of hardware appliances as part of software processing life cycle, in a similar way Analytics Hardware/Appliance as part of IoT  processing will become a key factor in 2015.Already  such  devices are available and  below are some of  the early implementers.

CISCO Fog Computing: Fog Computing is a paradigm that extends Cloud computing and services to the edge of the network. Similar to Cloud, Fog provides data, compute, storage, and application services to end-users. With the concept of  Fog  Computing, where  by   the  network locally analyze the IoT  data  and take a decision on what data to be passed on to cloud. It's a concept called fog computing. And Cisco® makes it possible today with the Cisco IOx platform. Cisco IOx takes the best of Cisco IOS® Software capabilities, combines them with compute, storage, and memory at the network edge.

SQL Server 2014 For Embedded Systems: Microsoft SQL Server 2014 for Embedded Systems provides a comprehensive database platform foundational for data analytics and operational intelligence in the enterprise. Microsoft SQL Server 2014 for Embedded Systems provides a comprehensive database platform foundational for data analytics and operational intelligence in the enterprise. With this combination  it  is  easy  to  make  purpose  built  analytics hardware   that  can  help  the  IoT  local decision  making. We  have not  seen any appliance from DELL, HP  etc.. target at the Fog Computing kind of embedded analytics for IoT. However with the  support  of   these  platforms these devices will come into the mainstream  shortly.

GE Proficy® Historian IPC: An integrated data collection and analytics appliance, Proficy Historian IPC delivers quick time-to-value for collecting real-time production and process information by simplifying purchase and installation.  Proficy® Historian IPC collects real-time production and process information for quick time to value; provides fully integrated simplicity of a purpose-built unit.

There  could  be  more  players  who  started  their  work  in this direction already.  We  may  conclude  that   Ánalytics  Hardware For IoT' may be  a  big thing  in  2015  along  with the  rest  of  software  framework  that  is already  part  of  it.` Also  the  exact  nature of Analytical capabilities needed for these devices are not fully specified as a standard across vendors, but with the initiatives like Fog Computing they may be standardized in the coming  days.

If you are a OEM and working on  Analytics  Hardware  for  IoT please write to me for adding to the list of vendors in the above list.

More Stories By Srinivasan Sundara Rajan

Highly passionate about utilizing Digital Technologies to enable next generation enterprise. Believes in enterprise transformation through the Natives (Cloud Native & Mobile Native).