Transcript
Of course, the obvious question is what's the use cases, how are we gonna use streaming, what sectors can benefit from streaming. And really, when you think about it, it's really more a question of what sectors, what businesses do not need streaming, do not need to act on real time data. And the real answer is probably very few.
Streaming Analytics Use Case: Brick and Mortar Retail
Just a few examples here, and it's not an exhaustive list, but for instance, say retail, brick and mortar retail. They have to compete with online, and the way to compete with online is to engage customers in the moment, whether in the neighborhood with location based marketing or when they walk into the store and are picked up by beacons. At that point, you have the customer while he or she can react and take action. You can engage them and get them involved and get them to make a purchasing decision.
Streaming Analytics Use Case: Manufacturing
Manufacturing is another use case, and again, this is actually not a--this is actually a fairly--it's not a new use case. We're monitoring the operation of machinery. We've done that for years, but typically, over relatively slow serial networks, we might monitor them, take readings maybe every few seconds.
Well, now, of course, we have machines that are so heavily instrumented that we have basically a huge amount of data coming in split second time, and the idea in manufacturing is that the one thing you really want to avoid is unplanned down time. And the idea is that if you basically can monitor all this data in real time as it's happening and perform these, basically these analytics, you can start to basically nip symptoms in the bud. You can catch some--when machines start to go off spec, and then you can plan around for maintaining them and plan around it before things break.
Streaming Analytics Use Case: The Connected Car
Another use case is the connected car. Now, of course, for a number of years, cars have become very data oriented. What's different today is that they're becoming connected as they're sharing data through wireless networks. And of course, we talked about self driving cars. They have to be connected. You have to be connected so that you can--so that a car can react in real time to basically, when there's all of a sudden a car in front of them that makes a sudden stop or when the light turns red, it's all about being connected, it's all about processing data in the moment.
Streaming Analytics Use Case: Cyber Security
Another good example, another good real-time example is cyber security. You want to catch those bad people, those bad actors before they wreak havoc and as opposed to an hour later when, oops, we've just had a denial of service attack, which has caused us to basically lose our defenses, and all of a sudden, 100,000 customer records have been compromised. You want to prevent that.
Streaming Analytics Use Case: Weather
Weather is another good case and actually has very good relevance today here in New York where we're getting a drenching rainstorm, they're predicting a couple inches of rain. Now of course, just like with manufacturing, very similar, we've had weather sensors, but typically, we've sampled them, maybe every five minutes, ten minutes, 15 minutes. Well, now that you have more ubiquitous networks and more sensors all over the place, we can get real time data.
And this is gonna be very crucial. For instance, here in New York, on a day like today where, basically, all of a sudden, we can catch where there's all of a sudden a downpour in one part of the city, the traffic department could be alerted to potential flooding problems. So, for instance, if you're out in the Central Plains where you can all of a sudden sense when a tornado is about to occur, again, having this data in real time provides better reaction time, which allows us to essentially brace ourselves, and it could become a public safety type of situation.
Streaming Analytics Use Case: Healthcare
And that might--very much translates to cases like smart cities, to healthcare, for instance - with healthcare, where you have high tech clinical devices monitoring patients.
Now, nurses and attendants can only --there's a finite amount of coverage that they can conduct, only a finite amount of new patients that they can monitor at any one time. But, if you have real time streaming analytics that are being fed from all these devices, basically, a professional can catch the signal--let's say a patient is about to stroke out--and then they can basically just like a machine that is starting to go off spec, they can catch this patient before the symptoms get really bad. And in this case, it becomes a matter of life and death.