Smart Homes at Scale
CIR Smart Homes & Cleanpower 2013 www.hvm-uk.com
pilgrim.beart@alertme.com
A	
  talk	
  in	
  3	
  parts	
  
•  Smart	
  Energy	
  Report	
  
•  Standards	
  and	
  Openness	
  
•  Home	
  Energy	
...
2013:	
  Quick	
  update	
  

…	
  
1.	
  Smart	
  Energy	
  Report
	
  
Data	
  analyCcs	
  engine	
  
•  An	
  engine	
  that	
  gives	
  insight	
  into	
  
household	
  usage	
  paFerns	
  fo...
MulC	
  Channel	
  CommunicaCon:	
  Reports	
  
• 
• 
• 
• 
	
  

Goes	
  out	
  with	
  the	
  bill	
  
Email	
  and	
  p...
Personal	
  Energy	
  Insight	
  &	
  Efficiency	
  Advice	
  
2.	
  Standards	
  and	
  Openness
	
  
Standards	
  and	
  Openness	
  
•  As	
  a	
  market	
  matures,	
  there	
  is	
  a	
  sudden	
  “flip”	
  
–  From	
  “n...
TSB	
  IoT	
  Interoperability	
  Demonstrator	
  
•  £6m	
  project,	
  1	
  year	
  
•  Goal:	
  Break	
  down	
  the	
 ...
Everyone’s	
  system	
  architecture	
  
You	
  need	
  data	
  from	
  several	
  services	
  
All	
  services	
  use	
  open	
  standards	
  

https:
But	
  each	
  is	
  organised	
  differently	
  
/customers/building/room/temperature

/users/hubs/devices/

/localauthori...
So	
  for	
  each	
  service	
  you	
  have	
  to…	
  
•  Read	
  the	
  documentaCon	
  
•  Write	
  code	
  specific	
  t...
Everyone	
  wants	
  an	
  ecosystem	
  

If	
  each	
  applicaCon	
  is	
  specific	
  to	
  each	
  service	
  
we	
  cal...
…but	
  Humans	
  Don’t	
  Scale	
  

But	
  adapCng	
  10	
  ApplicaCons	
  x	
  10	
  services	
  
=	
  100	
  pieces	
 ...
Problem:	
  
Services	
  not	
  machine-­‐browsable	
  
An	
  applicaCon	
  cannot	
  automaCcally	
  discover	
  a	
  new...
HyperCat:	
  
Makes	
  services	
  machine-­‐browsable	
  
What	
  have	
  
you	
  got?	
  

This:	
  
X/Y/Z	
  

re	
  
s...
HyperCat:	
  Easier	
  life	
  for	
  everyone	
  
•  Developers	
  

– More	
  data,	
  quicker	
  

•  Service	
  and	
 ...
HyperCat	
  is	
  not	
  a	
  panacea	
  
•  ApplicaCons	
  and	
  Services	
  sCll	
  have	
  to	
  agree	
  on	
  high	
...
Work	
  in	
  progress…	
  
All	
  the	
  things	
  we	
  kicked	
  out	
  of	
  scope!	
  
•  Data	
  formats	
  (SenML)	...
From	
  edge	
  to	
  centre	
  

They’re	
  the	
  same	
  thing.	
  
One	
  day,	
  almost	
  all	
  
Open	
  Data	
  wi...
“Open”	
  
Not	
  (necessarily):	
  
•  Free	
  
•  Public	
  
Means:	
  
•  My	
  service	
  works	
  with	
  your	
  ser...
We	
  may	
  think	
  we’re	
  “the”	
  service
	
  
But	
  we’re	
  just	
  another	
  node	
  in	
  the	
  graph
	
  
Smart	
  Meters:	
  T-­‐2	
  years	
  &	
  counCng	
  
Here	
  come:	
  
•  Consumer	
  Access	
  Devices	
  
•  Authorise...
3.	
  Home	
  Energy	
  Use	
  Update
	
  
Home	
  appliances	
  &	
  devices	
  

In	
  2009	
  the	
  average	
  household	
  owned	
  11	
  
Cmes	
  more	
  consu...
Electricity-­‐	
  per	
  Home	
  
0.00035	
  

0.0003	
  

0.00025	
  

0.0002	
  

0.00015	
  

COMPUTING	
  
ELECTRONICS...
Electricity	
  –	
  all	
  homes	
  
8,000	
  

1	
  
3	
  

7,000	
  

6,000	
  

5,000	
  

4,000	
  

GAS	
  
3,000	
  ...
(bubbles-­‐internet)	
  
(bubbles-­‐this	
  mac)	
  
Bill	
  =	
  Price	
  x	
  Usage	
  
•  Energy	
  prices	
  are	
  rising	
  
•  Household	
  bills	
  are	
  rising	
  
–...
@pilgrimbeart	
  
AlertMe Beart Smart Homes & Cleanpower 2013 Cambridge, UK via CIR www.hvm-uk.com
Upcoming SlideShare
Loading in …5
×

AlertMe Beart Smart Homes & Cleanpower 2013 Cambridge, UK via CIR www.hvm-uk.com

639 views

Published on

Smart Homes & Cleanpower 2013 Cambridge, UK via CIR
www.hvm-uk.com

Published in: Technology, Business
  • Be the first to comment

AlertMe Beart Smart Homes & Cleanpower 2013 Cambridge, UK via CIR www.hvm-uk.com

  1. 1. Smart Homes at Scale CIR Smart Homes & Cleanpower 2013 www.hvm-uk.com pilgrim.beart@alertme.com
  2. 2. A  talk  in  3  parts   •  Smart  Energy  Report   •  Standards  and  Openness   •  Home  Energy  Update  
  3. 3. 2013:  Quick  update   …  
  4. 4. 1.  Smart  Energy  Report  
  5. 5. Data  analyCcs  engine   •  An  engine  that  gives  insight  into   household  usage  paFerns  for  energy   consumpCon   –  Built  upon  several  models   •  Models  informed  by  many  inputs:   –  Temperature,  LocaCon,  Type  of   appliances  and  usage   –  and  others    …   •  Models  include:   –  Space  HeaCng,  Hot  water,  Cooking,   LighCng,  Appliances,  Occupancy   –  and  others  …   •  Output  improves  with  more  data   –  Supply  more  informaCon,   •  Get  greater  insights   –  Supply  less  informaCon   •  SCll  has  good  insight,  but  based  on   naConal/local  averages  
  6. 6. MulC  Channel  CommunicaCon:  Reports   •  •  •  •    Goes  out  with  the  bill   Email  and  paper  versions   Longer  term  trends  and  analysis  over  the   term  of  the  bill   Monthly  digests  
  7. 7. Personal  Energy  Insight  &  Efficiency  Advice  
  8. 8. 2.  Standards  and  Openness  
  9. 9. Standards  and  Openness   •  As  a  market  matures,  there  is  a  sudden  “flip”   –  From  “no-­‐one  to  be  open  to”   –  To  “openness  is  essenCal”   •  Can’t  do  it  all,  need  ecosystem  partners   •  Customers  want  to  have  choices  
  10. 10. TSB  IoT  Interoperability  Demonstrator   •  £6m  project,  1  year   •  Goal:  Break  down  the  verCcal  M2M  silos!   •  ~40  enCCes,  most  already  with  verCcal  end-­‐to-­‐end   plaborms   1248.io,  Aimes  Grid  Services,  AlertMe,  Amey,  ARM,  AvanC,  BalfourBeaFy,  BRE,   BriCsh  Telecom,  Carillion,  CriCcal  Sodware,  Ctrl-­‐Shid,  EDF,  Enlight,  ExplorerHQ,   Flexeye,  Guildford  Borough  Council,  IBM,  Intel,  Intellisense.io,  Intouch,   LivingPlanIT,  London  City  Airport,  Merseyside  Transport,  Milligan  Retail,  Neul,   Open  Data  InsCtute,  Placr,  SH&BA,  Stakeholder  Design,  Traak,  UK  Highways   Agency,  Westminster  City  Council,  Xively  and  the  UniversiCes  of  Birmingham,   Cambridge,  Lancaster,  Surrey,  UCL  &  Open  University   •  8  clusters  with  diverse  use-­‐cases:   –  Airports   –  Transport  logisCcs   –  Schools  and  Campuses   –  Streets   –  Homes  
  11. 11. Everyone’s  system  architecture  
  12. 12. You  need  data  from  several  services  
  13. 13. All  services  use  open  standards   https:
  14. 14. But  each  is  organised  differently   /customers/building/room/temperature /users/hubs/devices/ /localauthority/street/post
  15. 15. So  for  each  service  you  have  to…   •  Read  the  documentaCon   •  Write  code  specific  to  that  service  
  16. 16. Everyone  wants  an  ecosystem   If  each  applicaCon  is  specific  to  each  service   we  call  it  “verCcal-­‐integraCon”.     To  grow,  we  need  to  go  “horizontal”   and  build  an  ecosystem  where   all  applicaCons  work  with  all  services  
  17. 17. …but  Humans  Don’t  Scale   But  adapCng  10  ApplicaCons  x  10  services   =  100  pieces  of  code  to  write     (and  imagine  1,000,000  ApplicaCons…)  
  18. 18. Problem:   Services  not  machine-­‐browsable   An  applicaCon  cannot  automaCcally  discover  a  new   service’s  resources  …  so  a  human  has  to  write  code  every   Cme  to  enable  it  to  do  that.     I  don’t  know   where  to  start  
  19. 19. HyperCat:   Makes  services  machine-­‐browsable   What  have   you  got?   This:   X/Y/Z   re   structu  by   a   rowse b etadat by  m earch   s
  20. 20. HyperCat:  Easier  life  for  everyone   •  Developers   – More  data,  quicker   •  Service  and  Data  providers   – More  customers   •  End-­‐customers   – More  choice   •  Ecosystems  and  markets   – Removes  barriers  
  21. 21. HyperCat  is  not  a  panacea   •  ApplicaCons  and  Services  sCll  have  to  agree  on  high   level  semanCcs   –  i.e.  if  a  service  provides  temperatures  in  °C  then  the   applicaCon  needs  to  understand  °C   •  What  HyperCat  does  is  enable  an  applicaCon  to  find   those  things  that  it  does  understand,  in  any  service   –  e.g.  “show  me  all  the  resources  which  are  in  °C”  
  22. 22. Work  in  progress…   All  the  things  we  kicked  out  of  scope!   •  Data  formats  (SenML)   •  Ontologies  (general,  and  more  &  more  specific)   •  RegistraCon   •  Standard  Licenses   •  Key  management   •  MoneCsaCon  models  
  23. 23. From  edge  to  centre   They’re  the  same  thing.   One  day,  almost  all   Open  Data  will  come   from  IoT.   Billions  of     Cny  sensors   Very  large   databases  
  24. 24. “Open”   Not  (necessarily):   •  Free   •  Public   Means:   •  My  service  works  with  your  service   •  We  can  swap  providers  without  a  lot  of  effort   •  Requires  less  trust  
  25. 25. We  may  think  we’re  “the”  service  
  26. 26. But  we’re  just  another  node  in  the  graph  
  27. 27. Smart  Meters:  T-­‐2  years  &  counCng   Here  come:   •  Consumer  Access  Devices   •  Authorised  Third  ParCes  
  28. 28. 3.  Home  Energy  Use  Update  
  29. 29. Home  appliances  &  devices   In  2009  the  average  household  owned  11   Cmes  more  consumer  electronics  items   than  they  had  in  1970,  and  three  and  a   half  Cmes  more  than  in  1990     [Av  home  has  41  appliances  today]  
  30. 30. Electricity-­‐  per  Home   0.00035   0.0003   0.00025   0.0002   0.00015   COMPUTING   ELECTRONICS   LIGHT   COLD   0.0001   WET   COOK   0.00005   0  
  31. 31. Electricity  –  all  homes   8,000   1   3   7,000   6,000   5,000   4,000   GAS   3,000   2,000   COMPUTING   ELECTRONICS   LIGHT   COLD   WET   1,000   0   COOK  
  32. 32. (bubbles-­‐internet)   (bubbles-­‐this  mac)  
  33. 33. Bill  =  Price  x  Usage   •  Energy  prices  are  rising   •  Household  bills  are  rising   –  But  slower,  and  not  hugely  above  inflaCon   •  Because  per-­‐house  usage  is  falling   •  SCll  a  long  way  to  go  (up  &  therefore  down!)   •  PoliCcal  talk  of  banishing  green  taxes…   •  The  real  issues  remain:   –  InsulaCon   –  Gadgets  
  34. 34. @pilgrimbeart  

×