Turn IoT Device Data into Actionable Intelligence: How to Harness Information for Use in Your Data-Driven Business

Published: 01/March/2016

Reading time: 25 mins

Businesses run on data, and with the Internet of Things (IoT), the potential sources for that data has increased exponentially — along with the potential opportunities to capitalize on that information. The ability to quickly analyze data from an ever-increasing number of internet-enabled devices, before that information loses its value, is critical.

By positioning SAP HANA in an IoT edge environment enabled by Red Hat technology, companies can connect their core business processes to the edge of the network and gain instant access to a wide range of data, both online and offline. Plus, ensure the ability to collect and analyze massive quantities of information in real time.

During this live Q&A, Red Hat’s Prasanna Sivaramakrishnan and Sid Sipes of SAP discussed how to harness big data and leverage edge computing to get real-time insights that lead to quicker decision making. Readers asked data and IoT questions on security, data transformation, and device connectivity, such as:

How do I get data from an edge environment to SAP HANA?

What are the potential security concerns with IoT edge environments?

How can I handle the bandwidth of hundreds of sensors?

What is the best strategy for getting data into SAP HANA from my various devices?

What is the difference between embedded and distributed computing?

Meet the panelists:

Prasanna Sivaramakrishnan, IoT Solutions Architect, Red Hat
Prasanna has 11 years of experience providing enterprise software solutions to clients in the retail, insurance, banking, transportation, telecom, energy, and software industries. At Red Hat, he currently works for the IoT/Embedded Program out of San Francisco. He is focused on solving business challenges by developing solutions for device connectivity, data integration and management, and IoT platform augmentation.
Prasanna’s former roles involve positions in software development for enterprise systems and consulting roles for custom software applications utilizing middleware at companies such as HP, Oracle, and JP Morgan. He has an MS in computer science from the University of North Carolina at Charlotte.

Sid Sipes, Senior Director, Portfolio Solutions Group Go To Market, SAP
Sid joined SAP in 2010 via the Sybase acquisition and has over 25 years of experience in the database technology field, and in particular in the enterprise data warehouse and large data disciplines. Before the acquisition, Sid spent 15 years with Sybase, serving in data warehouse and financial services positions.
Sid is currently part of the SAP Digital Enterprise Platform Group and is responsible for the SAP IoT Edge Compute platform, as well as the SAP SQL Anywhere mobile and embedded database platform.

Transcript

Natalie Miller, SAPinsider: Hello! Welcome to today’s live Q&A on how to turn Internet of Things (IoT) device data into actionable intelligence. I’m Natalie Miller, features editor of SAPinsider and insiderPROFILES, and I’m excited to introduce today’s panelists, Prasanna Sivaramakrishnan of Red Hat and Sid Sipes of SAP.

Prasanna Sivaramakrishnan is an IoT solutions architect at Red Hat and has 11 years of experience providing enterprise software solutions to clients in the retail, insurance, banking, transportation, telecom, energy, and software industries.

Sid Sipes is senior director of Portfolio Solutions Group Go To Market for SAP and has over 25 years of experience in the database technology field and, in particular, in the enterprise data warehouse and large data disciplines.

Hi Prasanna and Sid, thank you so much for being here today to answer readers’ questions!

Prasanna Sivaramakrishnan: Hello, everyone! Thanks for joining this panel. I am excited to be here working alongside Sid from SAP to answer any questions you have about IoT!

Sid Sipes: Hello, everyone. I’m really happy to be here, and I’m looking forward to a good conversation. Thank you, Natalie, for organizing this session and providing us the opportunity to chat.

Natalie Miller: We are so happy to have you both here. Questions from readers are already starting to come in, but I’d like to kick things off with a question for you both. Where is the industry today in terms of development of Internet of Things technology?

Sid Sipes: Hi, Natalie. The industry is rapidly evolving as companies realize the tremendous value that utilizing and responding to IoT brings them. I think we will see many exciting developments in the very near future.

Prasanna Sivaramakrishnan: Great question! Currently we are looking at a deluge of devices – about 50 billion – to be used by enterprises and consumers in the near future. From a technological perspective, there is a need for Open Standards to handle the scalability, security, and reliability of these devices.

Natalie Miller: Great! Thanks for that answer.

Comment from Guest: What are the most promising sectors from an SAP HANA perspective in the multi-faceted catch-all called IoT?

Sid Sipes: IoT is a broad field that is promising benefits for virtually all industry sectors. We see benefits in acquiring, using, and responding to IoT data in retail, oil and gas, manufacturing, transportation, etc. Everything is connected and producing data, and this data has tremendous value — especially when combined with traditional business data.

Comment from Sundar: What are the values gained by doing IoT data analytics at the edge of the network? Kindly give me some examples.

Prasanna Sivaramakrishnan: Certain use cases in transportation, retail, or industrial automation require real-time alerts or feedback, which can drive actions. For example, in retail it can mean enhancing customer shopping experiences in a store and supporting them with sales reps and coupons targeted to their specific interests. Instead of relying on a backend cloud layer, a lightweight rules engine could easily provide the required results.

Sid Sipes: We also see many technical values gained. Some examples include: reduction of data flow from the devices to corporate, greater responsiveness gained by running analytics at the edge, better security, and the ability to run analytics and/or processes should connectivity with corporate go down.

Comment from Guest: Can you give examples of the adoption intensity of SAP HANA?

Sid Sipes: While I can’t give you specific numbers, I can tell you that HANA is the fastest growing product in SAP history. It has gone from 0 sales to over $1 billion in just a few years.

Prasanna Sivaramakrishnan: Additionally, there are integrations supported from Red Hat JBoss Middleware products, such as Fuse and data virtualization, that expand SAP HANA’s reach with open source technologies.

Comment from Buck: What is the difference in running analytics on edge versus running them centrally? Is running on edge specific to IoT?

Sid Sipes: Running on the edge is not specific to IoT. Think of it as running corporate reports on your tablet or mobile device. Another example might be field service personnel needing or wanting to run effectiveness reports while in the field (i.e. on an oil rig or in a factory).

The main difference in the analytics will be in the total amount and types of data available at the edge. Due to smaller devices with less storage, there will probably be a limited amount of data retained there

Prasanna Sivaramakrishnan: Running analytics on the edge is a broad term indeed. It could be similar to running on-premise software instead of relying on a third-party data center. With IoT, there is a need for lightweight analytics on edge that can then be supplemented with a predictive analytics approach governed by machine learning algorithms and, possibly, a big data implementation in the cloud or data center.

Comment from Buck: If we want to run analytics on edge, I am assuming we need some sort of database on the edge device. How does HANA come to reside on the edge?

Sid Sipes: The definition of an “edge” device is evolving somewhat, and it also varies by industry. For example, an edge device in transportation may be a small, single-board computer in a vehicle. An edge device for a factory floor or an oil rig may be a large server installation.

For those large installations we can see HANA residing in those server installs. In fact, SAP has an “edge” version of HANA available.

However, the vast majority of IoT edge devices are rather small, such as the single board computer I mentioned previously or the many, many deployments of network gateways and routers. For this type of deployment, SAP has a purpose-designed and built, embeddable database called SAP SQL Anywhere. It’s designed as a fully featured, ANSI compliant SQL DB that will run on small devices and in remote and unattended locations (no sys admin, no dbas).

Comment from Guest: How do I get data from an edge environment to SAP HANA?

Sid Sipes: There are many ways to get the edge data into HANA. For example, HANA has programmatic APIs for custom feeds. HANA also has a built-in stream engine designed to take high-speed data feeds. SAP also has a technology called RDSYNC, which is designed to automatically feed HANA from an embedded edge database. Traditional methods such as ETL and Replication are supported. This includes near real-time replication from the edge (RDSYNC).

Prasanna Sivaramakrishnan: With Red Hat’s partnership with SAP you can see certified and supported connectors from technologies such as JBoss Fuse, which can facilitate the data flow from the edge.

Comment from Buck: Are you able give more information on the “edge” version of HANA? For example, what are the minimum requirements for running the HANA Edge Edition?

Sid Sipes: The “edge” version’s official name is HANA Edge Edition. The best thing would be for you to do is a quick Google on it rather than for me to try to remember the specs from memory.

A quick note, however – the edge version is not just targeted at IoT Edge environments, but also at small-to-medium businesses. There are many use cases that it will be applicable for.

Comment from Guest: What are some best practices for handling the bandwidth of hundreds of sensors? What about situations where a device designed for one vertical domain needs to talk a device from another?

Prasanna Sivaramakrishnan: There are many best practices involved here. I will touch on three.

Choice of protocol to communicate between sensors and an IoT gateway/cloud platform: Protocols such as MQTT, AMQP, STOMP, and CoAP, which are based on Open Standards, are commonly used. These can be applied when devices from one vertical domain need to talk to other. Companies like Cisco are working on building 3G/4G networks to handle the bandwidth.
Choice of software to process the messages: JBoss A-MQ provides high scalability and reliability for processing the messages.
Choice of cloud/backend layer: You can leverage SAP HANA with predictive analytics or a big data strategy.

Sid Sipes: If you look at the tremendous growth in devices producing data, it becomes apparent that there is not enough bandwidth, so something has to be done. We think that providing the ability to accumulate, aggregate, and react to the data – at the edge – will tremendously help reduce the bandwidth issues. Not all of the device data will be needed at corporate (think about periodic health check pings). Aggregating and filtering data at the edge will allow the transmission of only what is necessary. Also, having the ability to store and process data at the edge means that processing at the edge can continue, unaffected, when the connectivity to corporate goes down.

Comment from Raman: How does HANA and IoT help the process automation industry in terms of continuous process improvement?

Sid Sipes: This is a question that is out of my area of expertise, so take this answer with a grain of salt. Using IoT data with HANA can help the process automation industry in the same way that it helps other industries.

You can combine IoT data with corporate business data. This allows the business side of the house to run processes and analytics in near real time and with real and current data. This provides the best results possible.
HANA includes a suite of predictive analytics, predictive maintenance, and other functionalities. We’ve used these functionalities in the past to monitor IoT data, look for trends, and provide feedback to the edge processes – either in terms of actions or, in some cases, new rules to operate at the edge.

The real-time monitoring and feedback loop can help any industry respond, react, and evolve much quicker than previously possible.

Comment from Guest: What are the benefits of integrating Red Hat technology into my HANA environment?

Prasanna Sivaramakrishnan: The benefits are that you get certified open source technologies following open source standards, which are beneficial when developers are building out mission critical or production grade applications.

Sid Sipes: Red Hat technology has been a vital part of the SAP landscape, including forming the underlying operating system and environment for HANA. We see that same relationship extending benefits to the IoT Edge. In addition, Red Hat has the ability to provide additional interfaces to sensors, devices, and buses over and above what SAP natively has.

Comment from Guest: How do we best handle security issues with data coming in from multiple devices?

Prasanna Sivaramakrishnan: Data coming in from multiple devices can be secured using some of the following techniques:

Use TLS (transport layer security) with a messaging infrastructure.
Use protocols such as WEP, WPA, or WP2 to secure the WIFI network.
Use techniques such as whitelisting and IP validation to make sure data from rogue devices are ignored.
Set up token validation or custom encryption to ensure a handshake between devices sending data to a software platform.

Comment from Guest: Is there a view to make HANA an Open Standard (Open Source Framework Standard Body)?

Sid Sipes: Not at this time that I know of.

Comment from Guest: For a company that is new to IoT, what should we consider to best prepare our middleware environment?

Prasanna Sivaramakrishnan: While designing your middleware environment, you want consider the following points:

Device connectivity using the right type of protocol.
Scaling and security of the middleware to handle the messages.
Usage of microservices or SOA-based implementation based on enterprise integration patterns.
Ability to scale the middleware using a SaaS model.
Integrations with disparate technologies, such as the Red Hat – SAP partnership, where you can leverage the platform and middleware technologies for developing mission-critical applications, in turn providing high ROI.

Sid Sipes: Evaluate what device data is absolutely required. This can help you shrink your bandwidth requirements. Also, look at contingency plans for network outages. Can your middleware environment (servers, gateways, routers, devices, etc.) help you recover and/or continue processing when outages occur? Between Red Hat and SAP, we can help you ensure that outages will not stop you or your processing. We can also help you with the architecture planning for both hardware and software.

Comment from Guest: Hi. I know an oil and gas company that wants to move to HANA and IoT for getting information from sensors. Can analog sensors be connected easily to HANA? Thanks.

Sid Sipes: Hello. Yes, this is one of the fields/industries in which we have quite a lot of experience in connecting sensor data directly into a HANA system. We have both native connectors for devices/sensors, which we have seen before, and for those who may be new to SAP, we have partners such as Red Hat that will have the sensor/device connectors.

Natalie Miller: As we come to the end of today’s Q&A, I’d like to thank you all again for joining us. And a special thank you to Prasanna and Sid for being here today and for all your insightful answers!

Sid Sipes: Thank you, Natalie, for giving us the opportunity to engage with people interested in IoT. I enjoyed this session. IoT is one of my favorite topics and is a technology that I think is about to change the world. Further information can be found at https://www.sap.com/iot. And, if you would like to continue the conversation, please don’t hesitate to reach out to me at sid.sipes@sap.com. Thanks again everyone! Sid.

Prasanna Sivaramakrishnan: Thanks everyone for joining! I’m looking forward to engaging with you guys for more exciting discussions on IoT in Collaboration with SAP. Feel free to drop me at question on my LinkedIn page or follow me on Twitter @ZenPras.

How do I get data from an edge environment to SAP HANA?

What are the potential security concerns with IoT edge environments?

How can I handle the bandwidth of hundreds of sensors?

What is the best strategy for getting data into SAP HANA from my various devices?

What is the difference between embedded and distributed computing?