The HDB Database involve collecting data from multiple sources and typing it in one Excel table, an original compilation Made by Teoalida. The database has been purchased by several companies, including home insurance, telecommunication, door-to-door services, real estate portals, flyer distributors, as well as many people who did not told me what business field they are in.
I offer 5 standard packages, containing block number, street address, building name, building type, building era, postal code, plus some specific details shown below. You can purchase a smaller package then pay the difference to a bigger package. For custom packages or additional data that is not yet included in database, contact me!
Instant download! Pay via PayPal or credit card and receive automatically an email with download link (example email).
FREE updates every 3 months for one year! Pay 50% per year for next years! See also Terms and conditions!
HDB Database statistics
Total number of tagged HDB buildings in database: 14789 (after Nov 2017 BTO update).
Database include also few hundred non-HDB buildings (not tagged, not included in this count).
Breakdown by building type tag:
Residential flats: 10715 + 1348 demolished (including blocks with shophouses and no flats).
Multi-storey carparks (MSCP): 1078 + 3 demolished.
Pavilions: 799 + 5 demolished.
Industrial buildings: 46.
Other buildings: 538 + 81 demolished (commercial buildings, electrical sub-stations, etc).
Unknown buildings: 45 (mostly demolished).
Breakdown by building era tag:
Classic era (lease pre-1993): 7827 buildings.
Centralized refuse chute era (lease 1992-1999): 2084 buildings.
Household shelter era (lease 1998-2007): 2028 buildings.
Household shelter BTO era (lease 2005-present): 2841 buildings.
Number of units is available for 10,506 blocks that were existing at least as 2010, including under construction blocks launched until November 2017 BTO. Total: 1,129,347 units, averaging 107.50 units per block. Every year, new blocks are completed while others are demolished, keeping this figure up-to-date is difficult. 1,017,335 dwelling units are in use as 31 March 2017 according HDB Annual Report.
Accuracy guaranteed for block numbers, street addresses, postal codes, lease year, number of units.
Not guaranteed for number of floors and units per floor, and for building type tag. Some 2-storey shophouses may be tagged incorrectly as residential. I need to find correct way to tag shophouses.
What the HDB Database currently include
– Block numbers: complete for all Singapore. Source: www.onemap.sg.
(40% done in 2009-2010, remaining 60% done in June 2014)
– Building name: complete for all blocks that have a name. Source: www.onemap.sg.
– Street address: complete except blocks under construction and blocks demolished before 2000. Source: Enquire status of upgrading programmes.
(30% done in 2010, remaining 70% done in December 2014)
– Town / Estate: complete for all Singapore. Town is official zoning while Estate is my personal classification by phases of construction.
– Town Council: complete for all Singapore. Source: www.onemap.sg.
(added in May 2017 at suggestion from a customer)
– Electoral division (GRC / SMC): complete for all Singapore. Source: www.onemap.sg.
(added in September 2017 at suggestion from a customer)
– Construction code & unit numbers: available for 3011 of 14586 total buildings, or 2464 of 2666 BTO era buildings.
In 2016 a customer suggested adding unit numbers. Source: site plans from BTO and SBF brochures, available for 90% of BTO era blocks and for random pre-BTO blocks which HDB offered site plans at Sale of Balance Flats.
I know how useful would be to have the unit number ranges for every block, but this is NOT POSSIBLE. Hong Kong does have www.centamap.com showing unit number and position. When Singapore is going to have the same? Going from block to block to check mailboxes and get unit numbers, takes more than a man’s lifetime.
– Building type tag: complete. This column act also as filter, non-HDB buildings have it left empty, who wants a simple list of HDB buildings can filter it to delete empty rows (estate header row, schools, condos, etc). Flats / MSCP / Pavilion / Other + Demolished tags, so the real estate agents can filter and delete demolished buildings, while historical researchers (like myself) can keep them.
(added in November 2014)
– Postal codes: available for 12966 buildings standing as 2015 or under construction. Source: HDB Centralized Map.
Personally not interested in postal codes, but in July 2014 a customer suggested adding them, a 2nd customer asked for postal codes in May 2015, so in June 2015 I added them. HDB Centralized Map added info boxes for under construction blocks in late 2016, allowing me to add in my database too at January 2017 update.
– GPS coordinates, latitude / longitude: complete for residential buildings.
Suggested by 1 people in 2013, too hard task to write coordinates manually, NOT in mood to do it, but DONE in 2015 thanks to Renata who bought my database and made a script to run in Google Maps API and add coordinates. In 2017 I met Sik Su Wei who made a script to get coordinates from HDB Centralized Map and I integrated them into my database using VLOOKUP function of Excel.
– Lease commence date: complete for all blocks available on resale market (so NO year for carparks, rental blocks, or blocks within MOP).
(30% done in 2010 and remaining 70% completed in December 2014, year only, source: HDB Resale Flat Prices.) Since 2015 I started to improve by adding month and day, sourcing from HDB Centralized Map.
– Estimated completion date: complete for blocks under construction or under MOP: source: BTO brochures. When blocks fulfill MOP I replace estimated completion date with lease date. Update frequency: when new BTO are launched.
Update frequency for above items: every 3 months when new BTO are launched. Also adding old demolished blocks when I discover them on hm.onemap.sg or National Archive photos. My ULTIMATE GOAL is to have in database all SIT/HDB blocks ever built.
– Building year: available for 505 buildings. HDB never publish actual building date, except when the block is selected for SERS. I have indicated the building year in database only for SERS blocks and for the blocks who were leased significantly later than their actual construction date. The building year is sourced from old newspapers seen at National Archive (if possible), otherwise estimated using old Street Directories.
– Demolition year: complete for all 1448 demolished buildings. Source: estimation using old Street Directories (1984, 1988, 1991, 1993, 1995, 1998) or Google Earth satellite imagery (after 2000).
– Number of units (text form) & upgrading programmes: complete for residential buildings. Source: Enquire status of upgrading programmes for blocks completed as 2015.
(25% done in 2010, remaining 75% ON HOLD waiting for a customer paying me…. been paid in May 2015 and completed in June 2015 (complete for all blocks that were standing to at least 2010).
– Number of units breakdown by flat type in separate columns: complete for all residential buildings (about 9500 buildings). Idea used since 2011 for List of BTO projects. In 2014 I planned to use it in the future for HDB Blocks Database too. I decided in June 2015 to breakdown original number of units beside current number of units which is varying over time.
(done in 2014 only for Jurong West for showing SAMPLE file. Breakdown done in all towns in June 2015). Since 2016 I add number of units when blocks are launched rather than when completed, source: site plans from BTO / SERS / SBF brochures.
Note: number of units is VARYING over time in blocks with 1/2/3-room built up to 1990, due to people who buy 2 flats and adjoin them.
Update frequency: I add number of units for new HDB blocks when new BTO are launched, but I will NOT offer updates for block built up to 1990, would take large amount of time to reached every block and benefits are insignificant. They will remain “as 2015”.
– Number of floors, units per floor, estimated number of units: 40% done in 2009 estimated from shadow length in satellite photos, before Google Streetview introduction, may have errors +/- 1-2 floors, additional 20% done accurately in 2010 by counting blocks in Google Streetview.
In 2014 I decided to complete database without number of floors because it was taking too much effort and I believed that it is no longer needed, precise number of units became available in HDB enquire status of upgrading programmes, no longer need to multiply floors with units per floor.
In 2015 a customer (Renata, see her comment) offered me a BONUS of $400 to complete number of floors for remaining 40% of blocks, but after doing half of job she told me to STOP THE JOB because her boss don’t want to pay.
I can resume the job and do remaining 20% if another customer is interested in paying a BONUS for completing number of floors. If do you want also to count the 40% estimated in 2009-2010, I can do if I have sufficient time, and you need to pay my time.
Virtual Singapore, a very detailed 3D map, was announced in 2015 to be ready in 2017. In August 2017 it was announced that will be ready in July 2018. This map will help me counting floors easier and accurately, so I won’t count any more floors on Google Streetview until Virtual Singapore will be launched for public.
What I can include or change in HDB Database
– Reorganize towns/precint structure – it’s important to know WHO are the potential customers and how they intend to use the data, to provide the RIGHT data structure.
– Remove number of floors – Floors can be only counted manually on Google StreetView thus I can save a lot of time by NOT adding number of floors.
– Remove condos and other private buildings from database, stick on HDB only.
– Make a separate database “Condo Directory”. DONE in October 2015!
– Make an API and sell access to database via monthly subscription instead of one-time payment. I don’t know how to make a subscription payment. Maybe some webdesigner with higher experience than mine can help?
– Anything else necessary? Your ideas are welcome!
History (short version)
I started compiling data voluntarily in 2009, purpose: personal research about housing density in selected estates and use the results in designing own apartments, I abandoned in 2011 after completing 40% because it required too much effort for a non-paid job.
Being a voluntary project done without any specific target, I changed my mind numerous times, making the database a MESS, table layout being different from town to town depending by the date each town was compiled.
In 2012 I published on website a SAMPLE with one town, as marketing experiment, to see who (if anyone) is interested. I was willing to re-compile entire database, this time as PAID service for professional use. I hoped people to tell their requirements or suggest changes in data structure, negotiate a price based by amount of effort needed to compile the data, then do the job… but I was facing with lack of feedback, LAZY customers do not inform me exactly what data are looking for!
Took 2 years to get a rough idea what people are looking for… and resumed working in July 2014, I done A LOT of changes in data structure, cleaned (most of) mess, completed whole Singapore in December 2014 in terms of block number, address and lease year. Number of units remained as done in 2011 (25% completed), the rest 75% will be done AFTER finding a customer willing to pay 50% before 50% after. The long-waited customer came in May 2015, work started, completing number of units for ALL 10,000+ blocks by June 2015! As customers requested, I added also postal codes in June 2015 and unit numbers in 2016. Feel free to request additional data!
Total: hundreds hours of work, and counting number of floors / units per floor require additional hours be to completed. Maybe too much time considering number of people willing to pay for such data.
Read more: history (full story) & changelog.