By Laurie Aznavoorian
In preparation for your next cocktail party I advise you bone up on ‘Big Data’, it is definitely the topic of the day. If you have not formulated an opinion as to whether it is good or bad, or worse have no idea what Big Data is, you had better educate yourself quick smart. Otherwise you will be the loser at the party left out in the cold, unable to communicate and labelled a Luddite. You’ll be sipping your beer in solitude, wondering whether you should have skipped the bean burrito at lunch. It will be pathetic.
Around you others will be deeply embroiled in conversations that follow what is now a well-travelled trajectory. It begins with an account of the marvels of technology followed by comparisons of smart phone applications. In time the conversation takes a turn toward the melancholy as party goers broach the inevitable topic of the loss of privacy that goes hand in hand with Big Data and then face the sickening realisation they have already given more information to the internet than they are comfortable with.
Over the past few months I have attended a number of workplace conferences in various countries and cities and not a single event excluded the topic of ‘Big Data’. Organisations like Johnson Controls, Unwired, IBM, Ripple Effect and Teecom are all offering advice, clever strategies and services to tap into the bottomless well of data organisations have at their fingertips, or are providing new tracking devices to collect a specific genre of data specific to the workplace e.g. information on our movement and use of the environment. The obvious intention there is improvements in workplace efficiency.
Beginning with Worktech 13 in Melbourne, the obligatory presentation from Phillip Ross at Unwired highlighted the new technologies that will change the way we work. In particular he noted advancements in near field communications, RFID tracking and BNS ‘building nervous systems’ that track real time performance of both the buildings we occupy and the people in them. These are all poised to make significant impacts to workplace effectiveness using Big Data.
Andrew Marshal of Johnson Controls has been on the conference circuit too, presenting their concept of ‘predictive analytics’. Marshal suggests that by using heat maps, studying workstyles and capturing data through meeting room, desk and personal sensors we will not only have the knowledge to dispel common workplace myths such as: spaces are used all the time, we are working in new ways – not just in different styles and that workplace design is a gimmick, but that we will also be able to derive real benefits by capturing a snapshot in time. Once this data is linked back to the buildings we occupy our workplaces will be more proactive and responsive to us as users.
Similarly David Marks from Teecom began his presentation at KA Connect 2013 in San Francisco by outlining five trends to take note of.
- Everything is connected
- Pervasive social media
- Big data is key
- New user interfaces
- Location or context
Marks was not the only one referring to ‘context’ at KA Connect, clearly this is a hot ticket item for the future. Incorporating context is touted to be a key feature in the next generation of the web evolution, providing greater meaning to the data we search for. He pointed out 70% of mobile phone use occurs indoors and if we capture that data it will allow us to create ‘social buildings’ that have the ability to deliver personalised information based on user context. Teecom’s product ‘Guide Dog’ is a start, by using existing tools such as e mail space users can find, reserve and navigate the workspace. Guide Dog will also collect data on how the space gets used in real time.
At yet another conference at the Massachusetts Institute of Technology (not one I attended), principal research scientist Andrew McAfee at the M.I.T. Centre for Digital Business, said Big Data was “the next big chapter of our business history” and that it will “relace ideas, paradigms, organisations and ways of thinking about the world.” This is a big claim, because we know that data on its own is not inherently meaningful, it takes clever people to derive meaning.
One of the problems with mining Big Data results from a type of modelling that originates in the sciences; it can lead to over simplification. We look for predictable behaviours in hopes they will repeat themselves according to the laws of physics, but many Big Data applications attempt to attach mathematical modelling to human behaviour, interests and preferences and one thing we know about humans is they are far from predictable.
This problem was highlighted recently in the US when all of the Web-browsing trails, sensor signals, GPS tracking and social networking messages, predictive algorithms, artificial intelligence software and data troves couldn’t keep the stock market from going into a tail spin when a hoax tweet claimed the white house was attacked and Obama injured. The computers made trades based on key words and phrases instigating a sell off.
On the other hand it was through similar Big Data mining that the Boston Marathon bombing suspects were caught. Using pictures from cell phones, portable video recorders, TV and surveillance cameras in public places investigators observed the crowd and identified the key suspects. Unfortunately the social news and entertainment website Reddit at the same time was using registered users content in the form of links or text to wrongly identified innocent people as potential suspects.
This highlights the limits and short comings of Big Data, it is not always right and not always used appropriately. While some like Craig Mundie, senior adviser at Microsoft and co-author of a position paper for The World Economic Forum believes “There’s no bad data, only bad uses of data” others like David Vladeck the former director of the Bureau of Consumer Protection at the Federal Trade Commission doesn’t buy the argument that data is innocuous until it is used improperly.
To demonstrate his point, a person may spend time searching online for deep fat fryers; they may be looking for a gift or researching a report for cooking school. But to a data miner tracking every click stream in the hunt could be read as an indication of an unhealthy eating habit. Using a data-based prediction this information could be later used to reject the person for health insurance or influence a potential employer.
Another challenge is models do not just predict, they create what scientist call a behavioural loop. We feed data in, it is collected by an algorithm that presents us with choices and those choices steer our behaviour.
With knowledge comes power and with power comes responsibility. How we use the data we collect will be an interesting topic for all of us in the future. For those of us in the workplace game consider the question posed by Gervais Tompkin at KA Connect 2013. Studies have shown that sitting all day is as unhealthy for us as smoking; therefore, what are we as workplace designers going to do about it. What is our moral responsibility?
Put that one in your pipe and smoke it.
KA Connect 2013 Conference, San Francisco – Presentations by David Marks Teecom and Gervais Tompkin Gensler.
Worktech 2013 Melbourne presentations by Andrew Marshall Johnson Controls
Big Op-Ed: Shifting Opinions On Surveillance Cameras. National Public Radio, Talk of the Nation – April 22, 2013
Lohr, Steve; Sure, Big Data Is Great. But So Is Intuition. The New York Times, December 29, 2012
Lohr, Steve; Big Data Is Opening Doors, but Maybe Too Many; The New York Times, March 23, 2013