Pragmatic solution for data-age


300 hours of video are uploaded to YouTube every minute! Almost 5 billion videos are watched on YouTube every single day. This is one example of the pace of data growth and overall size of data is piling into zetta-bytes faster. Where am I going with this is relevance of rapid analysis of data measured in hours or days rather than the stereotypical months of traditional data mining. The result is an opportunity to derive meaningful insights to business with more emphasis on predictive analytics.

Predictive analytics is nothing new. Predictive analytics is a way to identify the probability of future outcomes based upon historical data. For example, from customer perspective, companies can predict a likely lifetime customer value or the probability of either loyalty or churn. Let us look at few use cases to examine the relevance and importance of predictive analytics in the world of enormous data.

A fashion retailer story: Predictive analytics helped in analyzing the campaign spends and predicting incremental campaign impact of spend. This helped the retailer in understanding where not to spend: Having a close look predictive analytics helps in understanding the relationship between customer segments and the marketing campaigns being interacted with. This retailer predicted the probability of a particular channel influencing online or offline purchases within specific customer segments, and while this was enormously useful in understanding how to spend budgets targeting more personalized digital campaigns, it was equally insightful into identifying spend that wasn’t contributing to incremental value. The power of predictive analytics came in determining, should the retailer pay for this ad, or will a sale happen organically through another channel or communication that might cost nothing or next to nothing? This allowed the marketing team to choose the right channels to most effectively and efficiently reach different groups of prospects and customers, and second, it provided the information required to personalize by sending the right content and message to the right segments at a very granular level.

Preventing hospital readmissions: Hospitals are turning to predictive analytics as they began to feel the financial pinch of high 30-day readmission rates. Real-time EHR data analytics helps hospitals cut readmissions by five to seven percent. This demonstrates how predictive analytics in real-time can analyze EMRs data to automatically identify and target patients at the highest risk of readmission early in their initial hospitalization when there is a lot that can be done to improve and coordinate their care, so they will do well when they leave the hospital. Notably, Kaiser Permanente has been working to refine its readmissions algorithms in order to better understand which returns to the hospital are preventable and which are not, a crucial distinction for value-based reimbursements.

Lifetime value analysis of a subscribers of communication service providers: CSPs are lately realizing that not all customers are the same. It is important for CSPs to assign a quantifiable dollar value to each customer, in order to prioritize various sets of customers. The Lifetime Value of a subscriber provides the predicted yield from each customer over the customer life. This helps CSP in offering high priority customers loyalty bonuses, preferential treatment through personalized service, better credit norms etc.
We are living in a world of data overloading. Predictive analytics can be a true human partner from elections to sporting events to the stock market on what the future will bring. Predictive analytics elevate human kind from an educated guess to data backed decision.

Trends in Edge computing – Ease & secure access of Industrial data


In IIoT ecosystem, IT and OT meet at the edge. The edge is where data is sampled and collected from the environment by instrumented things or devices. Edge-Centric Architectures extend elastic compute, networking and storage across the cloud through to the edge of the network. The following are few of the trends for maintaining security and uptime with edge computing while streaming data securely.

  • Cloud offloading to address communication latency: Handling more processing at the network’s edge reduces latency from the device’s actions. Use cases that are highly time-sensitive and require immediate analysis of, or response to, the collected sensor data are, in general, unfeasible under cloud- centric IoT architectures, especially if the data are sent over long distances.
  • Encryption on endpoint to safeguard data security: By and large, sensitive and business-critical operational data are safer when encrypted adequately on the endpoint level. Unintelligent devices transmitting frequent and badly secured payloads to the cloud are generally more vulnerable to hacking and interception by unauthorized parties. Additionally, many enterprises may need to secure and control their machine data on the edge level for compliance reasons.
  • Sensor fusion: Combining data from different sources can improve accuracy. Data from two sensors is better than data from one. Data from lots of sensors is even better.
  • Sensor hubs: Developers increasingly experimenting with sensor hubs for industrial internet devices, which will be used to offload tasks from the application processor, cutting down on power consumption and improving battery life in the devices.
  • Analytics on the Edge leading to low cost-of-ownership and secured data: Reduce potentially huge cloud-computing costs (because of the sheer volume of 24/7 sensor data) by allowing “fog computing,” where the processing would be done right at the collection process combining real time analytics, with only the small amount of really relevant data being passed on to a central location.
  • Gateway-mediated edge connectivity and management architecture pattern: As the widespread acceptance of modern, open-field protocol standards has reduced the need for traditional gateways in the field, the IIoT has created a need for a new breed of intelligent gateways that unlock the full potential of interoperability among diverse real-world devices and industrial Internet systems

IoT enabling Manufacturing Platforms progressively deliver better Connected Cars


In the Digital age customers prefer experience over features. Aided by plethora of IoT technologies, the evolution of connected cars with a balance of features vis-à-vis experience is phenomenal from vehicle-to-infrastructure (V2I) to vehicle-to-vehicle (V2V) and now vehicle-to-everything (V2X) leveraging vehicle-to-cloud (V2C). In my view IoT technologies are enabling connected cars to deliver value from integrated ecosystem of industries including automotive, telecom, health, insurance and industrial automation. Manufacturing platforms are designing and delivering connected vehicles that are truly leveraging the aplenty Internet of Things technologies as described below.

  • V2I, V2V and V2X end point connectivity with mobile wi-fi offload (wi-fi hotspots, 802.11u, 3G/4G LTE/latest 5G), DSRC roadside infrastructure (802.11p), consumer network (femtocells, dealer hotspots) and energy service providers (charging stations etc.)
  • In-Vehicle-Infotainment enabled with intuitive human-machine-interface (HMI), advance driver assist solutions (ADAS), voice communications, around view monitor, and rear seat entertainment with 4G LTE speeds available via built-in mobile hotspot enables services such as Internet radio, video streaming, Web browsing and personalized music etc.
  • Public/Private/Enterprise cloud connectivity offering real-time navigation, weather forecast, traffic information and online route planning, audio & video streaming, health data updates and remote car monitoring
  • Telematics promoting passenger safety with smart SOS (e-Call), a wide range of security features that keep drivers connected and safe in the event of an emergency, including automatic crash notification, stolen vehicle tracking, and roadside assistance.
  • Leapfrog towards Autonomous Vehicles with advanced state of the art embedded sensors and actuators, contextual voice recognition, interconnectivity V2I/V2V, improved decision making algorithms beyond ADAS and efficient navigation technologies enabling hands-free driving with right balance of driver-in and driver-out of the loop scenarios
  • Vehicle advanced diagnostics and analytics enabling cost reduction across industrial value chain tying insurance telematics, remote diagnostics, and condition-based maintenance
  • Enabling modern day payments with electronic toll collection, parking reservation and payment

Evolution of ERP in IoT Revolution


Over the decades Enterprise Resource Planning (ERP) has become backbone of business corporations. In wake of intelligent factories and intelligent operations getting enabled by device proliferation and IoT, ERP is shifting to be more nimble and agile. I would like to table the following trends in IoT that are defining the evolving ERP space.

  1. Cloudification of ERP: The need from ERPs to provide mobile applications and cloud based software that allows employees to use the tools while away from the office. This trend switching ERP from classical licensing towards “as a service” model.
  2. Emergence of Industrial Internet or Industrie 4.0: In the era of IoT/IIoT the networked interconnection of things , which are often equipped with ubiquitous intelligence coupled with Smart Factory, not only represent new systems that need to be interconnected with the EPR software, but also potentially new business segments. Amidst fluid developments, ERP systems will play a defining role in the future of Machine to Machine connectivity which is blurring boundaries between virtual and physical systems. Hence IoT drive the evolution of the ERP software in the future by making ERP flexible, intelligent and real-time
  3. Componentization of ERP with APIs to provide cross functionality: Componentization of ERP is another important phenomena. The larger ERP systems should offer tone down versions of their core products. ERP can enable business agility as smaller and standalone products partner across platforms with APIs providing cross functionality between products.
  4. Need for real time data to enable analytics led business: As ERPs head towards IoT and cloud adoption, fast growing small and midsize businesses trying to keep up with the needs of consumers. The larger organizations are trying to take factory and supply chain decisions beyond four walls of the organization while their employees and decisions maker are on the move. So it is essential for ERP to connect unstructured data from devices with structured data in the business and must be able to process, analyse and show all this data in real time. This require ERP vendors to invest in high performance computing, pervasive connectivity, web services, and other trends” in order to stay alive in IoT era.

Top Trends in IoT for Smart Manufacturing


IoT is taking smart manufacturing a step forward enabling demand-driven ‘smart supply chains’. IoT systems are being built on top of a solid technology platform including cloud, analytics, big data and mobile. But to fully implement IoT systems for smart manufacturing, businesses need to establish an integrated fabric of devices, data, connections, processes and people. I was part of a panel discussion in identifying top IoT trends for smart manufacturing. You can download he outcome of panel discussion below.

IoT Protocols for Connected Homes


IoT Protocols for Connected homes:

The standards are evolving in a right pace towards a standard or protocol that will glue together and can turn a pile of cool gadgets into a system that runs your whole house for you. The leading providers working on standards unification include, Samsung’s SmartThings, Belkin’s WeMo, platforms of Lowe’s and Staples, and smart-home specialist Insteon has a line of hubs and devices. The above examples are w.r.t having a vendor or carrier decide which products can connect each other. Few newer platforms are designed to offer a broader selection of products that consumers can add on easily as a long-term solution to home connectivity. The following are network protocols enabling Home IoT.

  • Wi-Fi: The ubiquitous wireless system will remain at the heart of most home networks, but many small, battery-powered devices won’t talk to it directly because of size and power requirements.
  • Bluetooth: The familiar personal-area network tackles IoT with the power-efficient Bluetooth Smart (or Low Energy) version and is expected to add longer range and mesh capability in 2016.
  • ZigBee: A mesh network based on the IEEE 802.15.4 standard and widely used in low-power home devices.
  • IEEE 802.11ah: A version of Wi-Fi with lower power consumption, due for approval in 2016.
  • Z-Wave: A low-power mesh technology licensed by silicon maker Sigma Designs and used in a wide range of connected-home devices.
  • 6LoWPAN: An IPv6-only version of IEEE 802.15.4 mesh networking.
  • Thread: A protocol introduced in 2014 and based on 6LoWPAN, with added features for security, routing, setup and device wakeup.
  • ULE (Ultra Low Energy): A recently introduced low-power version of the DECT cordless-phone network technology.
  • APPLE HOMEKIT: While not a communications protocol, Apple HomeKit is actually a software framework, allowing developers to build smart home devices that will connect directly to the iPhone and iPad and be controlled by a dedicated app.
  • Vertical specific: IEEE 1905.1-2013 – “IEEE Standard for a Convergent Digital Home Network for Heterogeneous Technologies”

Fundamentally protocols for “Connected Homes” or broadly for IoT have to enact four common communications models – Device-to-Device, Device-to-Cloud, Device-to-Gateway, and Back-End Data-Sharing. Alongside enabling above communication models, “Connected Homes” or “Home IoT” standards and protocols should address two fundamental layers which are applications and network. Applications handles how devices interact & understand each other. Application layer is focusing on developing unifiers that certify products to work well on the same standard. Few key players in the application space include AllJoyn, HomeKit, ZigBee, OIC, Brillo and Weave, and Z-Wave. Secondly, Network determines how data travels through wired and wireless connections. The protocols for network include, Wi-fi, Bluetooth, ZigBee, IEEE 802.11ah, Z-Wave, ULE (Ultra Low Energy) etc.

Trends driving the Protocol development:

The key trends currently defining consumer protocols for IoT in connected homes possibly include few of the following.

  1. People trends: At a generic level, increasing desire by the consumer for convenience combined with growing desire by the consumer for safety and security. This is the primo motto of developing protocols for next gen consumers.
  2. Technological trends: The confluence of several technology and market trends is making it possible to interconnect more and smaller devices cheaply and easily:
    • Ubiquitous connectivity is key to enabling connected homes: Many homes now have a broadband connection promoting device connectivity. Low cost, high speed, pervasive network connectivity, especially through licensed and unlicensed wireless services and technology, makes almost everything “connectable’’.
    • Miniaturization is the trend to manufacture ever smaller mechanical, optical and electronic products and devices. Examples include miniaturization of mobile phones, computers and vehicle engine downsizing. Coupled with greater computing economics, this has fueled the advancement of small and inexpensive sensor devices, which drive many IoT applications.
    • Computing Economics is expediting home IoT adoption: Breakeven volumes of devices getting connecting are being attained with Moore’s law continuing to deliver greater computing power at lower price points and lower power consumption
    • Widespread adoption of IP–based networking: IP has become the dominant global standard for networking, providing a well–defined and widely implemented platform of software and tools that can be incorporated into a broad range of devices easily and inexpensively
  3. Data and privacy in connected homes: The connected home has been called the next frontier for ‘data analytics’. Already, a simple smart meter can report energy readings to a utility every few seconds, compared with a standard one, which is read either once a quarter, or whenever a user or meter reader records it. Multiply that by possibly hundreds of sensors and devices in the connected home of the future – which may only be a few years away – and there will be a data tsunami. But this smart home vision, and the Internet of Things (IoT) generally, has set alarm bells ringing about consumer privacy and it’s becoming apparent that first-class data privacy and security protocols are critical if we as consumers are going to accept it

Top 5 Trends in Automotive OEMs


As disruptive technologies are changing the landscape of the automotive industry, Original Equipment Manufacturers (OEM) are feeling the impact. New inventions in automobiles mean that OEMs have to adapt. And if they are able to adjust quickly, they may be poised to win big. I was part of an eight member industry expert panel who came together to delve into the trends and drivers that are defining the space.

You can learn more from the infographic here: