IoT and BigData

The impact on IoT and Big Data. I presented the this talk at http://globalbigdatabootcamp.com/

Transcript

1. IoT  And  Big  Data   Sanjay      -­‐Enterprise  Data  Architect  and  Analy:cs,  Arrayent  Inc.     @sabhub1    A  R  R  A  Y  E  N  T    
2. •  About  Sanjay   – Enterprise  Data  Architect  and  Analy:cs,  Arrayent   Inc.   – More  than  20  years  in  IT  and  mostly  dealing  with   some  form  of  data   – Worked  at  Apple,  Accenture,  WesternUnion   – Ac:ve  Open  Source  Junky  à  Download,  Compile,   Install,  Ask  Ques:ons  and  Help  Improve  it  
3. Agenda   •  IoT   •  Impact   •  Use  Cases   •  Data   •  Readiness   •  Conclusion  
4. What  is  IoT?   “The  Internet  of  Things  (IoT)  is  a  scenario  in   which  objects,  animals  or  people  are   provided  with  unique  iden>fies  and  the   ability  to  transfer  data  over  a  network   without  requiring  human-­‐to-­‐human  or   human-­‐to-­‐computer  interac>on”  
5. IoT?   Future  is  here,  you  like  it  or  not,  you  will  be  sucked  into  it  
6. We  are  all  I(di)oT’s?   •  We  called  the  Big  Box  the  Idiot  Box   •  Now  we  all  have  smaller  version  of   Big  Box  in  our  hands  and  pockets   •  We  are  glued  to  it  all  the  :me   •  We  want  more  of  it   •  And  we  had  enough  of  it   •  We  started  to  think  OOTB…    
7. What  We  Want  Next?   •  We  want  everything  around  us  connected.   •  We  want  to  control  things  around  us   •  We  are  asking  for  it   •  We  were  actually  doing  it.  
8. IoT  Is  Large  Space   All  Use  Cases  You  Can  Think  of  
9. The  Good,  The  Bad   •  Good  News  is  IoT  is  coming,  it  is  happening   •  Bad  News  is  IoT  is  coming  faster   •  As  Per  Gartner   – Internet  of  Things  Installed  Base  Will  Grow  to  26   Billion  Units  By  2020  
10. Value  Add   •  Distribu:on/Logis:cs   •  Brand  Adapta:on   •  Feature  Op:miza:on   •  Usage  Pa`ern   •  Cross  Sell  Opportuni:es   •  Power  Usage/Op:miza:on   •  Recall  Logis:cs/Service  Logis:cs   •  …  
11. Analy:cs  
12. Use  Cases   •  Retail  and  Logis:cs   – Tracking  of  goods  on  an  item-­‐level  a  feasible   business  case,  including  inventory  accuracy,   reduc:on  of  administra:ve  overhead,  automated   customer  check-­‐out  processes  and  a  reliable  an:-­‐ thea  system.   – In-­‐store  beacons  
13. Data  Life  Cycle  
14.   Devices     Sense   What  is   Happening   Decide   What  to  do   Build   Your   Context   Act     Quickly   &  Consistently   •  Context  Aware   •  Predic:ve  and   Rules  Driven   •  Con:nuous   real-­‐:me  at   scale   Real  Time  Ac:onable  Insights   Sanjay.sabnis@arrayent.com   Latency  Needs  are  Milliseconds  or  less  
15. Are  We  Ready?     •  Big  Data  is  like  you  have  never  seen   before   •  Gathering  data  from  previously   unexplored  areas.   •  Not  the  absence  of  data.   •  It  is  about  not  missing  on  data  that  you   really  need  
16. Use  Cases   •  Assisted  living   –  Cost  of  nursing  home  is  increasing.   –  Need  for  round  the  clock  monitoring  is  challenging   •   Smart  Ligh:ng   –  Op:mize  use  of  street  and  building  lights  based  on  current   condi:ons   •  Traffic  Monitoring   –  Monitoring  and  analyzing  traffic  pa`erns  to  reroute  drivers   •  Waste  Management   –  Op:mizing  waste  pickup  by  measuring  container  levels   •  Security  &  Emergency  Detec:on   –  Detec:ng  radia:on,  gases,  and  other  hazardous  condi:ons  in   real  :me.  
17. Test   •  Data  Centers  will  be  overwhelmed  by  Data  Deluge   –  Exis:ng  Data  Center  Capacity  will  be  put  to  test   •  A  new  focus  of  real  :me  analy:cs   –  Exis:ng  Real  Time  Components  will  have  to  rethink  what  actually  means  in  IoT   •  Data  Reten:on/Storage  reach  a  new  level   –  Cloud  Storage  Needs  will  surpass  what  we  have  now   •  Network  Bandwidth   –  New  inven:ons  are  needed   •  Privacy/Security   –  End  of  the  Day  Consumer  will  take  control   •  Skill  Set   –  We  need  lot  more  data  engineers  and  scien:sts   •  Data  Science   –  Need  “Data  Science  on  the  Wire”   •  Value  of  Data/Time  to  react   –  Window  is  Gefng  Shorter  
19. What  We  Do   Arrayent  PlaMorm  enables  trusted  consumer   brands  to  implement  connected  products  and   systems  
20. Home   15,000 ! Messages per Sec! ~200 Milliseconds! Round Trip Latency! How  we  are  Connected   Device  Virtualiza:on  
21. Arrayent  Data  Architecture   Real  Time  Replica:on   Data  Center  1   Data  Center  2   Data  Center  3   Batch  Analy:cs   Quick  Search/Query   Device  Status   Region  1   Region  2   Region  3   Using  Cassandra   •  Isolated  Work  Loads   •  Easy  to  Manage   •  Scalable   •  Mul:  DC-­‐Region   Deployment   •  Use  Network  Anycast  
22. Other  Components   •  Apache  Usergrid   – Using  it  as  RBAC   – It  can  do  lot  more  than  that.  It  is  fancy  data  store   running  on  Cassandra,  Highly  Scalable   •  ELK  Stack   – For  massive  log  pipelining,  Storage,  Indexing  and   pre`y  neat  UI.   – Used  in  NOC   •  Kaia     – For  persistent  distributed  pub/sub  
23. Landscape   Arrayent   Periodic  Table  of  IoT  
24. Thank    You   Rush  to  Home  Depot.     The  MyQ  Connected  Garage  from  Chamberlain,  supported  By  Arrayent