IOT - Devices and Data Powering Contextual Decisions

By Mark Gildersleeve, VP and head of Watson IoT and Decisions Platform, IBM

As happens with almost every major new technology platform, change happens slowly at first and then accelerates suddenly.  The Internet of Things  (IoT) has finally escaped the early hype and today delivers meaningful value across most sectors of the economy.

Three main factors  contribute  to higher value:

• Billions of connected  sensors, gathering  a  steady  stream  of raw information from consumer products, industrial electronics, transportation  networks, and supply chains. Today, we can collect a range of types of data at a physical granularity and a sample rate that was not possible even a few years ago.  This data can range from  vibration  and temperature  readings in a factory to soil hydrology and atmospheric pressure on a farm.

• Interoperability  of devices and systems. Broader coverage and greatly reduced costs of telecommunications enable greater connectivity across a range of settings and device types. Furthermore, innovations in cloud and edge computing allow the data from these devices to be more easily integrated and shared.

• Advances in analytics and machine learning. Platform investments  in  data  management,  analytics, and artificial intelligence allow for the creation of custom systems to analyze these massive amounts  of distributed  data. Importantly,  this  analysis can  drive action,  all the  way from  actuators at the edge of the network to optimized routing of trucking fleets. Local data has driven control  of devices and systems for decades. From the earliest days of supervisory control  and data acquisition (SCADA)  and machine - to - machine  (M2M)  technologies, the  industry  has  worked  to  use data  to  head  off  unplanned  ma chine down time by responding to changing conditions more quickly and by looking for patterns to plan maintenance schedules. We can extract more value when we aggregate and normalize data into a form that allows  for  broader  consideration. In one simple example, we can improve the quality of data coming from a temperature sensor in part by comparing its results with neighboring sensors.

"With greater deployment of sensors and actuators,the integration ofcollected data sets,and continued advance in analytics and machine learning, IoT will deliver even greater value"

In  the  world  of IoT,  outcomes  are improved by making better decisions with  better  data.  By ingesting  and correlating sensed data along with parameters such as environmental conditions we can use many kinds of analytical tools to better  understand what is happening, and through deeper  analysis  dig  into  how  we might effect change. The true inflec- tion point comes when the synthesis and analysis of data enables higher- order contextual decisions that drive concrete actions.  This process from data  collection, to  data  analysis, to decision, and finally action can drive real time systems that control equip ment, can drive scheduling of specific actions  (such  as preventive  maintenance). I can also drive planning and forecasting systems that use our improved  understanding of the  world to better forecast the consequences of changing conditions.

One  area where  IoT  data  is turned into action is with airplanes, one of the  most  complex IoT  devices out there.  Let’s  explore  air  turbulence to  bring  this to  life. There are sensors on  a plane that  measure when the airframe experiences turbulence. A message is sent to the operations center in real time informing dis- patchers that  a turbulence  event oc- curred. Analytics are applied on thatdata to predict where turbulence will be  occurring  in  the  future.  Then, an alert is sent to the planes that fly through  that  “turbulent   air  space” so that  flight  routes  can be  altered to avoid the hazard. The outcome is fewer injuries to flight attendants and passengers, lower maintenance costs, and a return on investment measured in months, not years.

Let’s look at a two other industries to illustrate the  possibilities. The retail industry is undergoing  truly massive change  as  commerce  shifts  online, consumers become more connected and  mobile,  and  advertisers  have more insight and ability to target promotions.  Retailers have access to a range of data sets including  retail receipts; footfall traffic in brick and mortar stores as measured by a range of sensors such as Wi-Fi, Bluetooth, and cameras; online browsing patterns, and competitive analysis. They can  use  this  information  to  power new analytics to improve their supply chain by predicting demand more accurately and to provide more target ed promotions  to  consumers across all shopping modes.

Transportation around the globe is now powered by the collection and analysis of distributed data sets to increase efficiency and safety, reduce emissions,and enable infrastructure planning. Public  transportation   re- lies on accurate system and passenger flow optimization  to  ensure crowds move fluidly and safely, while measurements of on-time performance, road and rail conditions, and traffic patterns can enhance multi modal routing for freight.

In summary, IoT has started to enable exciting change across a range of industries. Dramatic reductions in the cost of sensors, networking, and computing has allowed for the collection of more data than imaginable only years ago. But real value is only delivered when the aggregation  and analysis of those data driven action what ever form that action might take: a setting automatically triggered at a device on the network edge,a precisely targeted and hope fully welcome purchase offer to a customer,or the redesign of a major industrial machine. With greater deployment of sensors and actuators, the integration of collected data sets, and continued advance in analytics and machine learning, IoT will deliver even greater value in the months to come.

Don't Miss ( 1-5 of 20 )