Smartphone-Enabled Urban Solutions: Mobility-on-Demand and - - PowerPoint PPT Presentation

smartphone enabled urban solutions mobility on demand and
SMART_READER_LITE
LIVE PREVIEW

Smartphone-Enabled Urban Solutions: Mobility-on-Demand and - - PowerPoint PPT Presentation

Smartphone-Enabled Urban Solutions: Mobility-on-Demand and Mobility-on-Sale Shih-Fen Cheng Associate Professor of Information Systems Deputy Director, UNiCEN Corp


slide-1
SLIDE 1

Smartphone-­‑Enabled ¡Urban ¡Solutions: ¡ Mobility-­‑on-­‑Demand ¡and ¡Mobility-­‑on-­‑Sale

Shih-­‑Fen ¡Cheng

Associate ¡Professor ¡of ¡Information ¡Systems Deputy ¡Director, ¡UNiCEN Corp ¡Lab Singapore ¡Management ¡University

2016 ¡SMU ¡Logistics ¡& ¡Supply ¡Chain ¡Symposium: ¡Urban ¡Mobility ¡& ¡Circular ¡Economy, ¡Singapore, ¡March ¡7, ¡ ¡2016

slide-2
SLIDE 2
slide-3
SLIDE 3
slide-4
SLIDE 4

It’s ¡More ¡than ¡Just ¡Ride-­‑Hailing ¡App

Mobility ¡+ ¡Crowdsourcing

Household ¡ Chores Mobile ¡Audit/Research Crowdsourced Logistics

slide-5
SLIDE 5

It ¡All ¡Begins ¡With…

slide-6
SLIDE 6

The ¡Significance ¡of ¡the ¡iPhone ¡Era

  • Accelerometer ¡(2007)
  • A-­‑GPS ¡(iPhone ¡3G, ¡2008)
  • Digital ¡Compass ¡(iPhone ¡3GS, ¡2009)
  • Gyroscope ¡(iPhone ¡4, ¡2010)

* ¡Sensing ¡linear ¡acceleration

slide-7
SLIDE 7

The ¡Significance ¡of ¡the ¡iPhone ¡Era

  • Accelerometer ¡(2007)
  • A-­‑GPS ¡(iPhone ¡3G, ¡2008)
  • Digital ¡Compass ¡(iPhone ¡3GS, ¡2009)
  • Gyroscope ¡(iPhone ¡4, ¡2010)

* ¡Significantly ¡improve ¡location ¡accuracy

slide-8
SLIDE 8

The ¡Significance ¡of ¡the ¡iPhone ¡Era

  • Accelerometer ¡(2007)
  • A-­‑GPS ¡(iPhone ¡3G, ¡2008)
  • Digital ¡Compass ¡(iPhone ¡3GS, ¡2009)
  • Gyroscope ¡(iPhone ¡4, ¡2010)

* ¡Get ¡true ¡north

slide-9
SLIDE 9

The ¡Significance ¡of ¡the ¡iPhone ¡Era

  • Accelerometer ¡(2007)
  • A-­‑GPS ¡(iPhone ¡3G, ¡2008)
  • Digital ¡Compass ¡(iPhone ¡3GS, ¡2009)
  • Gyroscope ¡(iPhone ¡4, ¡2010)

* ¡Sensing ¡angular ¡acceleration

slide-10
SLIDE 10

Smartphone ¡Penetration

10 20 30 40 50 60 70 80 90 100

China Japan Korea Singapore UK USA France 2012 2013 2014 2015

Source: ¡Google ¡Consumer ¡Barometer

slide-11
SLIDE 11

The ¡Big ¡Promise ¡(and ¡the ¡Big ¡Gap)

Service ¡made ¡available ¡(now!) ¡as ¡you ¡need ¡it

  • If ¡enough ¡willing ¡workers ¡can ¡be ¡found
  • If ¡workers ¡know ¡where ¡to ¡situate ¡themselves
slide-12
SLIDE 12

The ¡Big ¡Promise ¡(and ¡the ¡Big ¡Gap)

Work ¡whenever ¡you ¡are ¡free

  • Yet ¡task ¡planning ¡might ¡consume ¡big ¡chunk ¡of ¡

that ¡free ¡time!

slide-13
SLIDE 13

The ¡Gap: ¡How ¡Tasks ¡Are ¡Assigned

  • (Proximity-­‑based) ¡Push

* ¡For ¡taxi ¡booking ¡service ¡and ¡most ¡ride-­‑hailing ¡App

slide-14
SLIDE 14

All ¡These ¡Wasted ¡Minutes

Typical ¡Weekdays

0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23

vacancy booking-­‑ratio

slide-15
SLIDE 15

The ¡Limit ¡of ¡Individuals

  • Human ¡drivers ¡generate ¡traces ¡that ¡closely ¡

resemble ¡traces ¡expected ¡from ¡a ¡myopic ¡ approach ¡(ok, ¡but ¡not ¡great).

5 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 0.5 1 1.5 2 2.5 3 3.5 4 4.5

Inbound Flow (%) Outbound Trips (%)

Hour: 07:00

y = 0.9107x + 0.1524 R² = 0.7044

5 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 0.5 1 1.5 2 2.5 3 3.5 4 4.5

Inbound Flow (%) Outbound Trips (%)

Hour: 19:00

y = . 8 4 5 2 x + . 2 5 2 R ² = . 7 7 3 9

slide-16
SLIDE 16

How ¡Can ¡Computation ¡Help?

  • From ¡the ¡data: ¡the ¡more ¡you ¡strategize, ¡the ¡better ¡you ¡perform.
  • With ¡sufficient ¡computation ¡efforts, ¡our ¡algorithm ¡can ¡do ¡reasoning ¡

with ¡infinite ¡depth.

  • And ¡extra ¡thinking ¡pays ¡off!
  • 1.68

1.77 1.85

slide-17
SLIDE 17

The ¡Gap: ¡How ¡Tasks ¡Are ¡Assigned

  • Pull
slide-18
SLIDE 18

All ¡These ¡Wasted ¡Minutes

For ¡“Pull-­‑based” ¡paradigm, ¡workers ¡suffer ¡from:

  • Having ¡to ¡browse ¡through ¡long ¡list ¡of ¡tasks.
slide-19
SLIDE 19

All ¡These ¡Wasted ¡Minutes

For ¡“Pull-­‑based” ¡paradigm, ¡workers ¡suffer ¡from:

  • Having ¡to ¡browse ¡through ¡hundreds ¡of ¡tasks.
  • Having ¡to ¡manually ¡construct ¡work ¡plans. ¡
slide-20
SLIDE 20

All ¡These ¡Wasted ¡Minutes

For ¡“Pull-­‑based” ¡paradigm, ¡workers ¡suffer ¡from:

  • Having ¡to ¡browse ¡through ¡hundreds ¡of ¡tasks.
  • Having ¡to ¡manually ¡construct ¡work ¡plans. ¡
  • Not ¡able ¡to ¡plan ¡for ¡the ¡future.

8:45am 9am 8am 8:40am 6:15pm 6pm 7:30pm 6:50pm 12pm 12:10pm 1:30pm 1:40pm

slide-21
SLIDE 21

TA$Ker: ¡A ¡Predictive ¡“Push” ¡Platform

Historical ¡(mostly ¡indoor) ¡ mobility ¡traces Home Work 3 2 4 1 Gym Home Work Super-­‑ market Task ¡Recommendation ¡Engine

slide-22
SLIDE 22

The ¡TA$Ker Experiment

  • More ¡than ¡1,000 ¡SMU ¡students ¡participated
  • More ¡than ¡30,000 ¡tasks ¡completed ¡within ¡2 ¡months
  • Super-­‑user ¡phenomena:

– 20% ¡users ¡performed ¡more ¡than ¡80% ¡of ¡tasks – Achieved ¡by ¡devoting ¡more ¡time – Efficiencies ¡are ¡lower ¡than ¡normal ¡user, ¡especially ¡when ¡in ¡ “Pull” ¡class ¡(25% ¡lower)

  • When ¡compared ¡against ¡“Pull”, ¡“Predictive ¡Push” ¡users:

– Finish ¡more ¡tasks ¡(56% ¡vs ¡44%) – Incur ¡less ¡detour ¡time ¡(50% ¡less)

slide-23
SLIDE 23

Some ¡Testimonies ¡(after ¡our ¡experiment ¡ended)

You ¡deprived ¡me ¡of ¡my ¡livelihood!! I ¡will ¡have ¡to ¡eat ¡instant ¡noodles ¡ everyday ¡until ¡TA$Ker opens ¡again!

slide-24
SLIDE 24

Takeaways

  • Mobility ¡+ ¡Crowd ¡=> ¡Mobile ¡Crowdsourcing
  • Workers, ¡part-­‑time ¡or ¡full-­‑time, ¡can ¡benefit ¡greatly ¡

by ¡having ¡customized ¡recommendations.

– By ¡incorporating ¡spatiotemporal ¡patterns ¡of ¡workers ¡and ¡ tasks.

  • Reducing ¡wasted ¡time/efforts, ¡not ¡exploiting ¡

participating ¡workers.

Follow ¡the ¡example ¡of ¡reCAPTCHA!