The Internet of Things holds great promise for everything from better healthcare to decreased traffic accidents to more efficient manufacturing processes. But these are still early days in the world of software, sensors, and network connectivity for devices, factories, vehicles, and buildings. Michael Morton is currently CTO at Dell Boomi, which he joined in 2013 after a career with IBM where he became an IBM Master Inventor and worked directly with Fortune 100 companies. Recently, Morton talked with BDQ about some of the opportunities and challenges and the role that Boomi plays in the emerging IoT market.
What does IoT mean from a data management perspective?
There are a lot of things to consider. One of the difficulties now is data overload. There is the question of how much data should be kept. And there are a couple of schools of thought on this. One is ‘tell me everything - I want all the data because then I am going to figure out how to be mine it for value.’ The other approach is to solve a specific problem and in this case, I would be selective about the data I want to retrieve and keep.
What does that mean?
Data management will be about how much data you need to keep and for what reason. This will lead to a whole host of interesting choices in terms of storage requirements, and cloud use, including network utilization and network bandwith considerations. Then, depending on whether you are trying to do analytics over a large sample of data in a data lake, the question is how fast do you need a result? Is it 5 minutes, or 30 seconds, or 2 seconds? That is really going to dictate your management strategy.
How does a data lake factor in as part of a data strategy?
The data lake is becoming a dumping ground for large quantities of data, structured data, and unstructured data in many forms. The data lake is really turning out to be a strategy where data is collected and then companies figure out how to mine data out of it.
How Dell Boomi fit into this picture?
There is a lot of opportunity for the integration platform as a service space. There is a lot of opportunity to leverage a solution like Boomi in really three primary ways. The first is data aggregation. Today, in the data lake there will not only be device data. When people think about IoT they think about data from a temperature sensor or airflow or a pipeline valve. But in fact, business value will come from not only device data, but also data from business endpoints. It could be customer information from a database, or some other historical information about customers or other types of business content. Business value will come from the aggregation of all this data. Boomi has a mechanism to be used to also aggregate business data along with device data in data lakes.
What else?
Another Boomi strength is data transformation. Customers use Boomi to access data from any number of end points they need and to transform, manipulate, and normalize data such as customer data from an on-premises system to a cloud platform such as Salesforce. Boomi can also take device information and transform it so that it makes sense before it gets written to say, a data lake, for example, transforming IoT sensor data from Celsius to Fahrenheit. Boomi provides a way to do transformations for interoperability.
And, the third?
The third strength is accessibility to data, without necessarily moving it when it just needs to be read or displayed in another application. Boomi provides a capability to web service-enable data from any source so if you have a database that has information and you have an application that needs to get data from that database, you can use Boomi as a mechanism to read data on the database, transform it, and then serve it back to that calling client as a web services interface.
Data aggregation, transformation, and web services enablement for data accessibility – which we call API management – are Boomi’s strengths. Atomsphere is the official name of the product that performs integration and interoperability functions, and then if you need web services enablement, you also select our API management solution.
So, IoT is a key focus for Dell Boomi but it is not the only focus.
IoT will be a new, evolutionary space for Boomi, which today is used for customers that are truly accessing, manipulating, and writing, typically, business content.
Where is the opportunity?
There are lots of solutions that are popping up to connect to devices, get data, and potentially talk to those devices, but the piece that is missing - that is the evolving opportunity for Dell Boomi - is how does a company know to contact a customer to tell them that there is something wrong, for example, that they need a new water filter for the water dispenser in their refrigerator?
This is the golden opportunity for Boomi because Boomi today is used for customer relationship management. Many customers use Boomi to integrate for example with Salesforce. So, if an appliance manufacturer was using Boomi, and the company received a message that my water dispenser needed a new filter, a Boomi process would take that information to “the last mile,” opening a ticket so the call center could reach out to me through an email or a call – and maybe they would also extend a coupon.
That is the beauty of IoT. There is one half of the world that is building solutions to manage connections and get device data, but there is the other half that ultimately involves an end customer or a consumer. Boomi is in a great positon to be that CRM bridge, to take that information, that event, and open a CRM ticket.
What does all this mean in terms of the data security challenge?
There are many aspects to this. In device security, risk comes from hacking into a device, a car, or a data warehouse. And this can involve not only the actual care of the product but also factor in terms of the distribution, being able to show the history or lineage of a product to prove, for example, that it has been stored in climate-controlled environment.
When it comes to device security, it is going to be very important to make sure that in a cold storage warehouse that no one hacks in to cause the loss of inventory, but also that if indeed someone did hack in, that there was a recovery mechanism that protected the inventory. You also need to be able to prove that you maintained the quality of your environment.
So device security has two aspects - there is the actual protection against a hack, but also making sure that you have no anomalies in your data records making it appear that you had a lapse in the quality of your environment when in fact you did not.
What are the other considerations for data privacy and IoT?
Today, I get an alert when my garage door is opened or closed. When I get notified, the information is going to my garage door company’s cloud so whenever my garage door opens, it sends an event to the company. I don’t know what they are doing with my data. If I had the door installed for 2 years, they would have data that shows every time I have opened and closed the garage door and if they were really crafty they could detect my living patterns and the likelihood that I am home or not.
Or, think of fitness trackers that have your fitness information, GPS data, and probably provide an indication of your health and how often you are exercising, or an air quality sensor that shows that someone is living an unhealthy lifestyle. Simple things that people don’t really think about can indicate the lifestyle of a person.
In a home situation, are consumers also able to access their own data?
These companies often provide interfaces to get at your data. Google provides a developer’s toolkit to get Nest thermostat data from their cloud. However, in the pages of terms and conditions, they limit you to only obtain and store 10 days of data. I don’t know why but you could speculate that the proprietary nature of Nest is to be an intelligent thermostat and they don’t want someone else building a better mousetrap.
You feel data governance is uncharted territory.
When someone agrees to use an Apple watch or garage door opener or Fitbit, who owns that data? There are a lot of unknown things going on now and the whole data space is wide open.
What’s on the horizon for IoT?
Many sensors cannot talk together today. Sensors talk to the manufacturer or the cloud provider but they don’t talk together. What you are going to see is more of a plug and play approach, with each sensor knowing that the other exists. There will be a normalization of cloud and plug and play through interoperability. My Google Nest cloud will talk to my wireless tags cloud to do data exchange and you will start seeing more evolution around common data models so devices from different vendors can talk to each other.
If you could give someone embarking on a big data project one piece of advice, what would it be?
“Incremental” is my key word. Go incremental. Start with a narrow problem. The challenge customers are having now is they have difficulty narrowing the scope of the first problem they want to solve because they are too busy thinking about the big picture and the what-ifs.
My advice is to just find the first problem you need to solve without trying to boil the ocean. Otherwise, you will spin your wheels because the market is changing so fast. But, you should also build a reference architecture.
Interview has been edited and condensed.
Image courtesy of Shutterstock.