1. 1 Make city roads smarter today! Pravin Varaiya Electrical Engineering & Computer Sciences, UC Berkeley Institute for Advanced Study, Hong Kong UST

2. 2 Outline  Why is traffic congestion so bad  Why is road network poorly managed  How to make our roads smarter

3. 3 Working population  Both % working age population and income are growing

4. 4 Vehicle growth faster than income  Vehicle ownership grows twice as rapidly as income in $3,000-$10,000 per capita range  Number of vehicles will triple by 2030 Fig Source: Joyce Dargay, Dermot Gately and Martin Sommer, 2007, Vehicle Ownership and Income Growth, Worldwide: 1960-2030

5. 5 Auto market in developing countries 14.47% annual vehicle growth in India

6. 6 Strains on road network  Population spread – Rapid spatial growth and urban sprawl  Under-resourced road network – Between 1970-2000 road network increased 3x, vehicles increased 16x – Peak-time average speed10-15 kph on main roads; 21-39 kph on arterial roads

7. 7 Why traffic congestion Increasing working population + Increasing income per capita + Decreasing car prices = Fastest growing car market in world Increasing car ownership + Increasing population spread + Under-resourced road network = Urban traffic congestion

8. 8 Cost of poorly managed roads  Effects of Congestion -Fuel loss at signalized intersections- Rs 995 crores per year, Rs 80 lakhs daily @ Moti Bagh crossing -Congestion delay costs Rs. 240 lakhs annually -Vehicles in congestion emit nearly four times more pollution -Autos greatest source of pollution (Delhi 66%, Mumbai 52%)  Accidents and fatalities –More than 25 accidents and 5 fatalities per day in Delhi –India has 1% of the world's cars, but 10% of traffic fatalities –5x fatalities of USA with 1/40x cars –Economic cost of accidents in Delhi is estimated to be Rs.35 lakhs per day  These costs can be reduced by better management Source: Delhi Traffic Police Dept, Economic Survey of Delhi 2002

9. 9 Why poor management  Transportation planners and engineers have many methods (theories and tools) to make road network efficient  These methods need comprehensive traffic data  India’s cities need urban traffic data system to support good road network management

10. 10 Urban traffic data system  Sensing, data backhaul, data base, applications:  Road performance  Signal control  Transit priority  Emissions reduction  Parking policy  Traveler information  Incident management  Diversion …

11. 11 Conclusion  Urban auto traffic demand will grow rapidly  In absence of intelligent management the cost in fuel, emissions, delay are excessive  Without measurement, management is blind  New (and existing) road networks should develop urban traffic data system

12. 12 Measure, measure, measure sensors access point

13. 13 10/26/2015 13 Slide 1 Area

14. 14 Carson St, LA  Performance based on measurement  Need analytics for multiple intersections Good Poor

15. 15 San Pablo Av, Albany  Answer requires analytics Why? 3-link 6-signal 0.9-mile median TT