When we met ParStream at MWC, we were handed a report from Dimensional Research that surveyed IoT stakeholders and their relationship with Big Data. The report was sponsored by ParStream, for reasons that will become apparent, and while there wasn’t time at the show to dive into the report, in the post-MWC glow, it still makes very interesting reading.
ParStream itself sells Big Data software, with its key product being the High Performance Compressed Index (HCPI), a patented approach that allows a user to index a much larger database in a way that allows real-time querying, response and analytics.
With big data, the speed of access is everything. For instance, knowing that a turbine requires unscheduled maintenance before it suffers an unexpected catastrophic failure, and therefore being able to shut it down before it explodes, can have an immense value to a business. In this sort of model, both the cost of replacing the turbine and the lost revenue through the downtime can be countered by predictive analysis. In situations like this, big data analysis is essentially an insurance policy – and insurance is big business.
A monthly license to use these services is an attractive proposition for businesses working at this scale – where real-time insight can prevent expensive repair bills. Businesses like ParStream are proving that there is money to be made in the sector, but these kinds of real-time and actionable insights are simply not valuable for the majority of businesses.
The retail sector has need for data analysis and collection, but it currently (and almost certainly never will) has no need for real-time analysis of shopping trends as it doesn’t have a mechanism to respond quick enough to that information – unlike a locomotive or wind turbine.
At MWC, ParStream CMO Syed Hoda said that the business was seeing increased demand from the industrial sector, and noted that many industrial facilities required better services due to their remote locations inhibiting the speed of an internet connection. At MWC, ParStream was showing off its new EdgeBoxAnalytics unit, which moves the HPCI functionality closer to the data source itself, so that network latency is greatly reduced.
So as the machine sensors get cheaper and the number-crunching hardware becomes more affordable to deploy, the Dimensional Research report says that “the cost to store data has dropped dramatically, while capabilities for analysis have made huge leaps forward. But are these connected devices actually gathering usable data?”
The key findings of the report are that there is a wide variation in IoT projects and their intended purposes, data challenges are not delivering the full potential of the IoT, and better data collection and analysis would provide more value. This is fairly commonsense stuff, but the numbers are very useful and deserve closer attention.
Of all the respondents, 53% are using IoT projects to optimize their existing business, with the remaining 47% treating them as strategic investments in future plans – perhaps a strange skew given the initial assumption that many readers would make that the majority of IoT projects would be more akin to moonshots than streamlining existing processes. The target audience for these projects were consumers (42%), businesses (54%), and internal use (51%).
Some 86% of stakeholders in business roles said that data was important to their IoT project, but only 8% said they are fully capturing and analyzing data in a timely fashion. This is due to 94% of respondents saying they faced challenges in collection and analysis – despite 70% of respondents saying that the data could enable more meaningful decisions, and 86% saying it would increase the return on investment (ROI) of their IoT investments.
Less than half of respondents said that their projects were aimed at consumers, which reflects Hoda’s view that the real money in IoT was in the B2B sector rather than the B2C market.
When asked what aspects of the IoT project have been most challenging; 58% answered business processes or policies such as privacy, 51% said user adoption of new technology, 41% identified timely collection and analysis of data, 40% said the sensors or devices themselves, and 4% said they had not faced any challenges with the project.
In terms of current data capture; 8% fully capture and analyze data in a timely fashion, 58% said they captured data but want to do more than their current level of analysis, 17% said they captured and stored data but don’t analyze it, and another 17% don’t capture data at all.
Of the 17% that don’t capture and data, most “simply were not far enough along the path to be capturing data yet, but they did have plans to implement capabilities.” Of the companies that do store data, 47% store data for more than one year, 24% store it for less than a year, 18% store for less than a month, 7% less than a week, and 4% for less than one day.
When asked what challenges are encountered in collecting and analyzing data, a few particular answers stand out. A full 44% said that there was too much data to analyze effectively, and we suspect that ParStream’s sales team will be trying to track them down. Another 36% said that it was difficult to capture data in the first place, with another 25% saying data was not captured reliably, and 19% saying that data was captured too slowly to be useful.
Once data is captured, 27% said they weren’t sure what to use it for and were unsure what questions to ask. Much like data capture, 26% said that the analysis process was too slow to be actionable, but a further 24% said that business processes were too rigid to allow them to act on their findings – even if they were captured and crunched in time to be useful.
A rather smug 6% of respondents said they faced no challenges in capture and analysis, but “interestingly, while cost is often a limiting factor in many technology decisions, for IoT stakeholders, ease of use appears to be a more pressing issue than cost. More participants (76%) say they would collect and store more data if it were easier than those who said they would collect and store additional data if it were free.”