Treatment on day to day applications
One of the futuristic technology trends we talk (also hear or read) about nowadays is about the Internet of things (IoT). We are bombarded by the big industry analysts and pundits predicting that by the year 2020 we would have at least 50 million devices connected with each other. The whole world is eagerly awaiting for the phenomenon when virtually all devices would communicate with each other and the concept of “dumb device” would cease to exist. While there is on one side – the much talked & hyped Consumer IoT there is the other angle (less discussed and known) - the Industrial aspect.
The Industrial Internet of Things (IIOT) has been prevalent for many years. The Data Centre industry is a big consumer of IoT and also a service provider for IOT. It is imperative that the Data centre business demands that it should have the smartest buildings and all IOT solutions deployed in those buildings. Thousands and thousands of sensors are communicating with each other on automated basis and there is no manual intervention at all.
A typical building of Data centre like the one we have in Mumbai is as single building at a footprint of maybe 300 thousand square feet and a consumer of around 30 MW of power. In a normal scenario that amount of power can feed a big locality. Essentially it means that the building houses thousands and thousands of IT devices running the most critical applications for either a bank or ecommerce service provider or a manufacturing company. The mandate is that these critical applications cannot go down for a single second because a single second of downtime can create a havoc to a bank, ATM, ecommerce site, Credit rating agency or broker site amongst others
The deployment of IOP starts from getting a power from substation grids - if the grid fails even for a second the diesel generator (DG) set should start, one DG fails the second should start the second fails the third should start. If the power load is so high that 1 DG set cannot take it, all the DG should start immediately. Similarly we take the entire component from the main grid to the panels to UPS right up to the rack level - at every level not only do we have the redundancy of equipment but we ensure that each of this equipment is supposed to be intelligent connected to the central Building Management Solutions (BMS) which is like IOT implementation that anything failing anywhere in the system - or maybe if the load in the beginning is increasing or either at par or from the cooling point of view the central analytics is able to drive the other equipments to kick start. This ensures that redundancy is never lost and any equipment is never going down to ensure the guaranteed uptime.
The Datacentre has thousands of sensors in the building and every sensor placed in front of every single rack which is continuously monitoring how much power is consumed and how much heat is generated by each of those devices for each of those racks. Depending upon the temperature profile in that particular area and by dint of analytics and the IoT solutions deployed we can create a heat map of the server hall at any moment of time. We then know which area is green, which area is blue, which area is yellow or red. Based on that the sensors can trigger which cooling unit to stop or start based on the load factor to ensure that the optimal use of power and make the system super-efficient.
Let us take another dimension – say the fire security solutions. Right from the fire detectors across the building on the roof to below the raised floors or on the rack – there are sensors all around. The moment two or three detectors detect an iota of fire or smoke being there – then immediately automated connected suppression system kicks in and immediately will throw in as much gas to make sure that the fire gets suppressed there and then.
So these are a few examples (out of the numerous) which illustrates the use of deployed IoT solutions in the different facets of Data Centre business which helps to run the operations effectively and efficiently. This is critical from the point of view from operational efficiency and cost efficiency.