I was reading some online articles and came across Technology Trends for 2014 - the number one being the 'Internet of Things' or IoT for short. This isn't exactly a new concept, - the promise of smart homes, where everything from intelligent lights to A.I. for washing machines that can be monitored remotely have been around for a while. And how couldn't resist the concept of a smart fridge that can stock itself? The term Internet of Things has been around for over decade, being firstly proposed by Kevin Ashton whilst Auto-ID Center at MIT, primarily driven by an interest in RFID, but the ideas and uses cases for an Internet of Everything has taken a while to mature.
The drivers behind the IoT have been several: The demand for renewable energy means that smart grids have to be able to monitor and respond to demand in electricity generation in a more agile manner - allowing for bi-directional energy supply from small energy producers (potentially you and me) - requires smart metering and monitoring. There's the exponential growth of smart phones - more people are always connected, and that trend will continue.
However, when the IoT was first envisaged all those years ago, there were some technology inhibitors:
1. The limitation of IPv4 in terms of the number of physical addresses that were available
2. The capacity of the internet for a fully connected IoT
3. The ability for mediators to scale to millions of concurrent connections
4. The ability to store and analyse the data in a scalable way
5. The ability to analyse all the data to make sensible decisions in a timely manner.
Fast forward to today and we have most of these things from a technology perspective either solved (e.g. IPV6) or the pieces are available, and Red Hat is ideally placed to provide the whole solution for a scalable backend for the IoT - and to do it all on open source software.
Firstly, we need the ability to provide a standards based, horizontally scalable solution for handling connectivity to hundreds of thousands of concurrent connections. JBoss A-MQ is combining the best of Apache licensed middleware solutions from Apache ActiveMQ, QPid and HornetQ to form a highly scalable messaging solution that supports MQTT, AMQP, WebSockets and STOMP.
The IoT will generate a lot of unstructured data, which needs to be correlated and analysed, and one of the leading NoSQL solutions for doing this is Hadoop. If you want Hadoop to scale and perform, then bester infrastructure to run it on is a combination of GlusterFS and OpenStack.
Getting real time data into Hadoop's HDFS can be problematic, but JBoss Fuse already has some of the best solutions for doing just that.
Finally, if you want to use complex event processing, to make decisions based on the flow of data from your connected devices based on causality and temporal logic, then JBoss BRMS is the best open source solution on the market.
Red Hat is going to be right at the centre of IoT solutions in 2014.
The drivers behind the IoT have been several: The demand for renewable energy means that smart grids have to be able to monitor and respond to demand in electricity generation in a more agile manner - allowing for bi-directional energy supply from small energy producers (potentially you and me) - requires smart metering and monitoring. There's the exponential growth of smart phones - more people are always connected, and that trend will continue.
However, when the IoT was first envisaged all those years ago, there were some technology inhibitors:
1. The limitation of IPv4 in terms of the number of physical addresses that were available
2. The capacity of the internet for a fully connected IoT
3. The ability for mediators to scale to millions of concurrent connections
4. The ability to store and analyse the data in a scalable way
5. The ability to analyse all the data to make sensible decisions in a timely manner.
Fast forward to today and we have most of these things from a technology perspective either solved (e.g. IPV6) or the pieces are available, and Red Hat is ideally placed to provide the whole solution for a scalable backend for the IoT - and to do it all on open source software.
Firstly, we need the ability to provide a standards based, horizontally scalable solution for handling connectivity to hundreds of thousands of concurrent connections. JBoss A-MQ is combining the best of Apache licensed middleware solutions from Apache ActiveMQ, QPid and HornetQ to form a highly scalable messaging solution that supports MQTT, AMQP, WebSockets and STOMP.
The IoT will generate a lot of unstructured data, which needs to be correlated and analysed, and one of the leading NoSQL solutions for doing this is Hadoop. If you want Hadoop to scale and perform, then bester infrastructure to run it on is a combination of GlusterFS and OpenStack.
Getting real time data into Hadoop's HDFS can be problematic, but JBoss Fuse already has some of the best solutions for doing just that.
Finally, if you want to use complex event processing, to make decisions based on the flow of data from your connected devices based on causality and temporal logic, then JBoss BRMS is the best open source solution on the market.
Red Hat is going to be right at the centre of IoT solutions in 2014.
