The Internet of Things’ impact on data centres
Mon 11 Jun 2018 | Leo Craig
In the final instalment of our mini-series about the hidden impact the Internet of Things has on data centres, Riello UPS General Manager Leo Craig rounds up some of the slightly unusual uses of interconnected devices you probably aren’t familiar with.
Over the last few weeks, we’ve highlighted some of the major areas where ongoing advances in interconnectivity and the Internet of Things (IoT) are leading to huge changes in our professional and personal lives.
In manufacturing we’ve learned how robotics, automation, and machine learning are influencing the ‘smart’ factories of the future. Industry 4.0 has potential to boost ‘UK PLC’ by hundreds of billions of pounds and stimulate significant job creation. However, with a typical smart factory producing upwards of 5 petabytes (PB) of data per week, there are huge capacity and processing questions for the data centre to answer.
Similarly in healthcare, we discovered how IoT and artificial intelligence promises to help ease the intense pressures our NHS faces to deliver both frontline care (i.e. robots assisting with surgery or AI offering diagnoses to non-emergency patients) and reduce the burden of administration for hard-pressed doctors and nurses.
In professional sport, we saw how data analytics is transforming everything from player recruitment, injury recovery, and personalised fitness training programmes, through to enhancing fans’ experience and influencing the flourishing fantasy sports and gambling industries.
While last week we delved into the $20 billion a year world of online video gaming, and how the popularity of platforms such as PlayStation Network and Xbox Live is likely to see monthly gaming data use top 3,000 PB by 2020.
Increased data capacity demands in those four areas pose their own specific challenges for data centre operators. But there are a whole host of other aspects of our day-to-day lives that have potential to be transformed by big data. You’re probably well-aware of concepts such as driverless vehicles, virtual assistants, or robot warehouse workers, but here are a few uses you might not so familiar with…
Unexpected Internet Of Things applications
The Long (Robotic) Arm Of The Law
Although there are natural concerns regarding civil liberties and privacy, AI is increasingly being used to fight crime. Last year, police in Durham started using a predictive intelligence algorithm that advised whether a suspect should remain in custody or not. The Harm Assessment Risk Tool (HART) forecasts the risk of reoffending using 34 variables including prior history of criminal behaviour, age, and where a person lives.
Earlier this year, the then Home Secretary Amber Rudd argued for greater use of data analytics to help identify stabbing hotspots, with machine learning identifying all crime reports that mention terms such as knives or blades. While counter-terrorism chiefs are using algorithms that have a 94% success rate at detecting terrorist propaganda and offensive material posted online and on social media.
Robots Aiding Recruitment
Blue-chip companies including Pepsi, Ikea, and L’Oreal are all employing AI technology to help recruit their next wave of employees. The firms are amongst the users signed up to ‘Vera’, a Russian-developed robot that uses AI to interview and sift prospective job candidates.
It can ‘interrogate’ as many as 1,5000 potential candidates during a nine-hour working day, something it would take human recruiters months to match. The software also scans CVs on job sites and automatically contacts candidates that meet the required qualifications to see if the person is interested.
Vera uses speech recognition and self-learning software that analyses 13 billion words. As well as playing a role in the hiring of workers, the robot can also conduct exit interviews with staff that are leaving a company.
Farming, Food Production, and Supermarket Fraud Detection
Many of the traditionally manual tasks required in agriculture and food production are now automated. ‘Wine-bots’ prune plants in vineyards, while ‘lettuce-bots’ ensure potentially damaging weeds are removed from crops. Robots can even operate ‘driverless tractors’.
Precision farming, which uses sensors and drones to measure and predict conditions such as water levels, estimated yields, and soil acidity, is helping farmers plan more effectively and increase crop harvests.
While once the finished product is on the supermarket or shop shelves, AI still has a crucial role to play. Online store Ocado improved the detection of fake purchases by a factor of 15 after implementing what it claimed was the world’s first AI-based fraud detection system for grocery orders.
Money Matters – AI In Banking And Insurance
In banking, predictive analytics software monitors transaction details in real-time, identifying spending habits and sending customers alerts about their account or even prompts to help control their spending. Similarly, many of the major online betting platforms use machine learning to examine betting patterns and flag up potential problem gamblers.
One of the biggest beneficiaries of the Internet of Things is the insurance industry. Sensors and GPS trackers allow for usage-defined car insurance calculated on real-time driving data. While advanced AI systems interpret medical certificates, surgical data, and hospital stays to automate the process of determining insurance pay-outs. A Japanese broker, Fukoku Mutual Life Insurance, replaced its 30 plus workforce with computers and increased productivity by 30% as a result.
What does the Internet of Things mean for data centres?
So we’ve explored a few of the more unusual applications where interconnected devices and sensors are transforming our lives. Of course, there are countless other areas the IoT is helping to transform too that we haven’t mentioned.
It’s really no surprise that by 2025 it’s predicted the average person will interact with a connected device 4,800 times a day – that’s once every 18 seconds! But what are the implications of all this additional data? It poses tricky questions for both the data centre and critical power protection sectors.
Previously, data centres were primarily a means for storage of information. But to really reap the rewards of AI and machine learning, data need to be analysed and processed virtually in real-time. This requires far greater capacity and power. And it puts additional strain on the critical power protection provided by infrastructure such as uninterruptible power supplies (UPS).
So many sectors rely on the Internet of Things now that the electricity going down for any length of time at all is unthinkable. Not only do such incidents cause catastrophic disruption and hit productivity, they pose the genuine threat of disastrous data loss. It makes a reliable UPS system and backup power supply a crucial element of any organisation’s risk management strategy and contingency plans.
While cloud storage capacity has increased significantly over recent years, the need for real-time processing has also seen the rise of edge computing. Having micro data centres on-site or as close to the location as possible offers far quicker processing times as the data doesn’t need to travel all the way to the cloud and back.
Modular data centres have undoubtedly played a crucial role in this trend, and they offer operators the flexibility to quickly scale up as and when they need it simply by adding in the necessary racks. The development of modular UPS allows for power protection to be delivered in similar ‘agile’, incremental fashion.
Compact and power-packed, modular UPS units deliver higher density power using less energy in a smaller footprint, so they are an ideal solution for hard-pressed ICT and facilities managers. Modules can be closely matched to the data centre’s power requirements and deliver the N+1 redundancy that is so crucial to minimise the risk of any damaging downtime.
The way data centres themselves operate is also being shaped by interconnectivity and automation. Many of the tasks previously carried out manually are now carried out automatically by computers or software. That means facilities are often practically unmanned, so it poses the question as to whether they actually need heating or lighting.
It’s clear the IoT, AI, and machine learning are on an unstoppable journey to transform practically every aspect of our lives. But it’s only through the unsung work going on behind the scenes at data centres housing all the petabytes of information, and the UPSs that ensure the electricity keeps constantly flowing, that billions of people around the world will continue to benefit.