Big Data: Where Data, Machines and the Individual Collide
Impacts to the Rental and Construction Industries
I was sitting in a round table discussion about big data at the European Rental Association convention in Stockholm earlier this month. The topic on the table was big data, how it can be effectively used within the rental industry to deliver benefit to both the rental business and rental customers. The discussion was fascinating, not least because 2 main themes developed even before the topic was really addressed.
First, how much confusion there is between big data and normal “reporting and analytics”. The latter being something that’s been practiced by most large businesses for a long time and the former being the new, somewhat scary, kid on the block. Second, how much fear there was around the table about security and privacy with a number of participants. They felt limited in what they could do with big data by the potential risk of privacy infringements or security concerns.
If we take the first of those themes as a starting point -- what is the difference between big data and ‘normal’ analytics? Is it just about data volume or is there another difference? Reality is that definition of “big data” is relatively fluid; however, three things that really differentiate it from traditional analytics are: volume, velocity and variety.
Traditional analytics generally comes from ERP systems and from transactional data. It tends to be generated on a relatively steady pace and though volume can be very large. If you are simply concentrating on each order placed and not; for example, on the path to the order that the consumer takes then the number of data points will be relatively small. Once you start tracking data based on website behaviour, start capturing location data from your customer’s phones or tracking Facebook likes and tweets, Instagram views then not only does the volume of data rise astronomically quickly, but the variety of data sources changes.
When it comes to security and privacy there is no doubt that the change in the variety of sources, the velocity and volume of incoming data can impact security and privacy. However, in meaningful terms, there is no difference in the security and privacy concerns around managing a traditional transactional database for old-style analytics which has your customer’s name, address, buying habits and financial information in it to managing incoming data from Facebook and machines. Whatever you were doing before to manage that data you still need to carry on doing, its just now you have more types of data and more data sources to consider. Those security and privacy concerns shouldn’t be enough to stifle your ability to use the information, there is another aspect to privacy however.
When it comes to the construction and rental industries there are obvious points of benefit when it comes to big data: more effective predictive maintenance, better machine design to reflect actual usage and improved safety. We are at a point where individual usage of a machine by an operator can be tracked against many data points. We can effectively track how safely the machine is being used, how much the engine is being pushed, how often the machine is being used at the edge of its safe working height / load / performance, how often the operator ignores a warning light, how many hours he or she worked without a break. This is no longer just about the machine, we could be tracking this against the individual. We could know that Fred behaves differently than Jack who behaves differently than Sarah and most importantly we can identify the potential effects of those different behaviours.
If 88% of all job site accidents are caused by human error (a statistic cited at the ERA convention) isn’t it our responsibility to capture and use this data at a human level not just a machine level? Big data gives us the capability to stop an accident before it happens, to recall an engine for maintenance before it breaks down. But in taking those decisions we inevitably conflict with individuals and their rights and with companies and their rights.
What do you do if the data shows that operator X is consistently working on the edge or over the safety limits, that operator X is more likely to have an accident than operator Y because he or she simply works too fast? What do you do if as a rental company you can see that operators on one construction site are consistently working too many hours for safe working or using your equipment dangerously? Who do you tell? What is the impact of that information on the individual or the company?
This is a discussion that needs to start happening more openly very soon. The technology exists, the machines exist to monitor and validate every activity, but where individuals fit into the story is still not entirely clear.
Source: Helen Sowerby, Director of Business Development for Wynne Systems - www.wynnesystems.com