RTA Future Move Hackathon, with Dataiku

One minute. Then the real day, in six numbers.Skip to the day

RTA Future Move Hackathon, with Dataiku

Insiyab.

The timetable said 19:23. The bus left at 19:36. Ride down to see why.

01 · The day

1,056,343

card taps on Dubai's buses in one day

11,201 trips1,071 buses166 routes

Every bus and every stop, on Monday 15 December 2025. This is the day we studied.

Measured on the real day

02 · The gap

47%

of buses leave late at 7 in the evening

16% at 7 in the morning3 times more in the evening

Same buses, same drivers, same timetable. Late means 5 minutes or more. Only the hour changed.

Measured on the real day

03 · One bus, one evening

+29min

The trip took 96 minutes. The timetable gave it 67.

67 in the timetable96 on the road16 break before the next trip

Route 310, 6 in the evening. The bus arrives 29 minutes late. Its 16-minute break is gone, so its next trip leaves 13 minutes late.

Illustration built from measured figures

04 · The delay travels

870of 936

late departures: the bus was late back from its last trip

93% came after a late arrival62% the bus was not back yet

Of 936 late departures we could trace, 870 came after the same bus finished its last trip behind time. By a stricter count, in 589 of 945 the bus was still out when its next trip was due.

Measured on the real day

05 · The time that is missing

244hours

of driving time the evening needs and the timetable does not give

630 hours across the whole day118 routes short of buses at 6 pm

Added up over every route, from 4 to 7 in the evening. Each pillar is a route; red means too few buses at 6 pm by RTA's own planning rule.

Measured on the real day

06 · The forecast

15%closer

Our forecast of the trip time is closer than the timetable, from 16:00 to 19:00

8.25 min off: timetable7.42 simple rule6.97 our forecast

Made 30 minutes before the bus leaves. Tested on 2,043 evening-peak trips it had never seen. After 19:00 the timetable is as close or closer.

Measured on the real day
Scroll to ride down
3D scene illustrative · figures measured

The problem

The timetable said 19:23.
The bus left at 19:36.

Someone at the stop planned their evening around 19:23. The driver did nothing wrong. The bus was still finishing its last trip, because that trip took 96 minutes and the timetable gave it 67.

The timetable gives each route the same trip time all day. In the evening the trip takes longer, so buses fall behind, and stay behind.

Illustration: clock times worked out from route 310's measured figures at 18:00

1 · It is not more passengers

48,724 vs 49,707

People boarding at 8 am and at 6 pm. The same.

But late buses go from 19% to 45%.

2 · The trip takes longer

67 → 96

Minutes in the timetable, minutes on the road.

Route 310 at 6 pm, a typical trip.

3 · The break is too short

16

Minutes between trips, the same all day.

Not enough to win back 29.

4 · So the next trip starts late

870 of 936

Late departures that came after the bus finished its last trip behind time.

In 589 of 945 it was still out when the next trip was due.

Which routes, at which hours, need more time, and how many minutes?

The four figures: measured on the real day

Why it is not the crowd

Same crowd.
Different clock.

The two peaks carry almost the same number of passengers. The morning copes. The evening does not. So the cause is not demand: it is how long the trips take against what the timetable allows.

boardings in the 08:00 hour
of departures 5+ minutes late at 08:00
boardings in the 18:00 hour
of departures 5+ minutes late at 18:00
Boardings in the hourDepartures 5+ minutes late, per cent

Read from the Dataiku flow: demand_by_hour and punctuality_by_hour. Hours with fewer than 500 departures are left off the line. Times after midnight belong to the same service day.

Measured on the real day

Where the timetable is short

hours missing.
In three hours of the evening.

For each route and hour we compare two numbers: the trip time in the timetable, and the trip time that would be enough for 85 trips out of 100. Add up the difference and you get the driving time the timetable is missing. Across the whole day it is bus hours.

Read from the Dataiku flow: running_time_ledger_by_hour. A second analysis by a teammate, with other code, finds 569 bus hours for the day.

Measured on the real day

Our answer

Forecast the real evening.
Tell the planner where to give it time.

A forecast

30 min

before each trip: how long will it really take?

15% closer than the timetable, 16:00 to 19:00, on an evening it never saw.

A worklist

244

hours of missing driving time, as minutes per route and hour.

Measured from the day, not predicted. The planner accepts or sets aside each row.

A second ranking

695

boardings in one hour on route C01 at 18:00, with 85% of departures late.

Where lateness reaches the most passengers.

What a planner gets

A list.
Not a dashboard.

The routes and hours where the plan falls furthest short, and the minutes each one needs. The planner accepts or sets aside each row. Nothing applies itself.

Read from the Dataiku flow: running_time_bands, rows with 5 or more trips, ranked by extra minutes times trips. In the Dataiku web app this is the "Planner's worklist" tab, with Accept and Set aside on each row.

Measured on the real day

The forecast

Know the evening
30 minutes early.

Gradient-boosted trees, trained on the trips before 16:00 and tested on an evening they had never seen. Average miss in minutes; lower is better.

8.25 min
RTA's timetable
7.42 min
Best simple rule: this route earlier that day
6.97 min
The model: 15% better, range 12% to 19%

Read from the Dataiku flow: running_time_metrics, trips due from 16:00 to 18:59. MAE is the average miss; RMSE weights large misses more; MAPE is the miss as a share of the trip.

Where it helps, hour by hour.

RTA's timetableThe model

What it does not do.

Only 6% past the simple rule.

Most of the gain comes from knowing how the route ran earlier that day.

Live signals add nothing at 30 minutes.

A bus's previous trip has not finished when the forecast is made.

After 21:00, the timetable wins.

It is an evening-peak tool, and we say so.

One day of data.

Every number comes from Monday 15 December 2025.

One more thing the day shows

The double-deckers
run closest to time.

Share of departures 5+ minutes late, by type of bus. This is an association, not a cause: the routes each type runs are not the same, and that is not controlled for here.

Read from the Dataiku flow: punctuality_by_bus_type, types with 2,000 or more departures.

Measured on the real day

How it's built

Inside Dataiku.
Where RTA can run it.

Eighteen steps in five zones: twelve visual recipes and six short Python recipes. One rebuild from RTA's raw tables takes 131 seconds. The web app's tabs carry the same names as the zones.

Zone 1

Passenger demand

Who boards, where, and when.

  • count_taps
  • demand_by_hour
  • demand_by_route_hour
  • demand_by_stop

Zone 2

Service reliability

Does the bus leave on time?

  • clean_running_time
  • punctuality_by_hour
  • punctuality_by_route_hour
  • punctuality_by_bus_type

Zone 3

Supply and capacity

Were enough buses sent?

  • buses_needed_rule
  • supply_status_by_hour

Zone 4

Forecast

How long will each trip take?

  • build_trip_table
  • forecast_running_time
  • running_time_bands
  • running_time_ledger_by_hour
  • train_running_time_model

Zone 5

Priorities

Where does lateness reach the most passengers?

  • passenger_impact_by_route_hour
  • passenger_impact_scored
  • passenger_impact_top25

Who uses it

A planner.
Every timetable change.

RTA changes timetables every one to two months and routes four times a year. Each night adds a day of records. Before each change, a planner opens the worklist, sees where the evening runs long, and decides how much time to give it. Boardings, never people "waiting". Nothing about one card, one driver or one vehicle is shown.

Give the evening
its time.

About 244 bus hours of extra running time from 16:00 to 18:59 would cover 85 trips in 100. About 630 across the whole day.

Now see it live in Dataiku

Opens the planner's web app on the real day. It asks for a Dataiku sign-in.