Industrial Project (236504) Transportation task planning algorithms ClickSoftware Project Requirements Students: Noam Lavie, Ori Shalev Supervisors: Israel.

Industrial Project (236504) Transportation task

planningalgorithms

ClickSoftwareProject Requirements

Students: Noam Lavie, Ori ShalevSupervisors: Israel Beniaminy, Simon Arazi

Introduction

• The operation of a modern transportation systems is extremely complex and it requires coordination between many factors:

– Regulator

– Time table

– Facilities: stations, platforms, ticketing self service kiosks, vending machines

– Network: railways, electricity supply, signaling, communication

– Vehicles

– People: drivers, call center, ticketing, passenger agents, maintenance, control room

• In this project we are focusing on train drivers and there working arrangement.

• There are many possible working arrangement, but some are better then others.

• The optimal arrangement covers the required service level (time table) with the minimal cost (headcount)

Our Objective

• To developed and test working arrangement algorithms for train drivers.

• The algorithm receives the train schedule (time table), and output a working arrangement for the drivers.

• The algorithm should find a valid working arrangement with the minimal headcount.

Time Table

• Explanation• Service ID: 1Y11LB, from Edinburgh to Fort Williams (Scotland)• This service is not active on Sunday, active Mon-Sat (0111111 = Sunday 0, Monday

1, Tuesday 1, Wednesday 1, Thursday 1, Friday 1, Saturday 1) • Departure from Edinburgh 04:50 arrival to Fort Williams 09:45. • Service duration is 5 hours and 4 minutes. Total distance is about 250 Km.

Train

ID

Days

Run

Route

Origin

Route

Destination

Start

Time Duration

End

Time

Route

Distance

1B01LA 1000000 FRTWLM EDINBUR 19:00 5:15 0:15 234.6049

1B01LB 0111110 FRTWLM EDINBUR 19:50 5:00 0:50 234.6049

1Y11LB 0111111 EDINBUR FRTWLM 4:50 5:04 9:54 249.3159

Service 1Y11LB Details

• Explanation• The travel between 2 consecutive station is a Leg (there are no stops within the leg).• Leg start is the departure time from the station, leg duration is the travel time to the

next stop. • Leg 1745A ends in 8:40, the next leg (1746A) starts at 8:45, it means that the train

stops 5 minutes in RANNOCH station.

Leg ID From Location To Location

Leg

Start

Leg

Duration Leg End Leg Distance

1739A EDINBUR WESTRTN 4:50 1:06 5:56 94.4381

1740A WESTRTN DALMUIR 5:57 0:06 6:03 15.04965

1741A DALMUIR ARCHRAT 6:05 1:02 7:07 58.80237

1742A ARCHRAT ARDLUI 7:08 0:14 7:22 7.975319

1743A ARDLUI CRNLRCH 7:22 0:20 7:42 8.750171

1744A CRNLRCH BOORCHY 7:45 0:28 8:13 12.56229

1745A BOORCHY RANNOCH 8:15 0:25 8:40 15.60018

1746A RANNOCH FRTWLM 8:45 1:09 9:54 36.13779

Scope

• schedule anonymous drivers to all legs (no fallouts).• The idea is to build a predefined valid working days (diagram), the collection

of all working days (diagrams) should cover the whole time table. • The goal – cover the time table with the minimal number of drivers

(diagrams)• The diagrams should be aligned with the following guidance:

– The driver starts his working day at a certain station, he must be there at least 10 minutes before the train arrival (leg start)

– The working day should not include more than 12 hours and not more than 250 Km driving

– It is preferred if the driver starts and ends his working day at the same station– The driver must have at least 1 hour break during the day, at least 30 consecutive

minutes– A driver can take a ride as a passenger with a train, or take a taxi to the next station

or walk to the next station

Example – Diagram ID GQ01007

ON GLGQHL   5:18 0:10 5:28  

TAXI GLGQHL DALMUIR 5:33 0:25 5:58  

REL DALMUIR   6:04 0:00 6:04  

DRIVE LEG DALMUIR ARCHRAT 6:05 1:02 7:071Y11LB

DRIVE LEG ARCHRAT ARDLUI 7:08 0:14 7:221Y11LB

DRIVE LEG ARDLUI CRNLRCH 7:22 0:20 7:421Y11LB

DRIVE LEG CRNLRCH BOORCHY 7:45 0:28 8:131Y11LB

DRIVE LEG BOORCHY RANNOCH 8:15 0:25 8:401Y11LB

REL RANNOCH   8:40 0:00 8:40  

PASSENGER RANNOCH BOORCHY 8:42 0:18 9:001Y42LA

PASSENGER BOORCHY CRNLRCH 9:02 0:28 9:301Y42LA

DRIVE LEG CRNLRCH ARDLUI 9:36 0:15 9:511Y22LW

DRIVE LEG ARDLUI ARCHRAT 9:52 0:13 10:051Y22LW

DRIVE LEG ARCHRAT DMBRTNC 10:07 0:51 10:581Y22LW

DRIVE LEG DMBRTNC DALMUIR 10:59 0:08 11:071Y22LW

REST DALMUIR   11:08 0:50 11:58  

DRIVE LEG DALMUIR HYNDLND 12:06 0:08 12:142H56LA

DRIVE LEG HYNDLND PTCK 12:15 0:01 12:162H56LA

DRIVE LEG PTCK CHRNGXG 12:17 0:04 12:212H56LA

DRIVE LEG CHRNGXG GLGQLL 12:22 0:02 12:242H56LA

WALK GLGQLL GLGQHL 12:24 0:10 12:34  

NOTE GLGQHL   12:34 0:00 12:34  

OFF GLGQHL   14:19 0:00 14:19  

Schedule – the diagrams cover the time table

In this example Driver 1 takes the train from DALMUIR to BOORCHY (Diagram ID GQQ0107C) and Driver 2 completes the journey (Diagram ID FW0951C)

Diagram

IDActivity

ID Type

From

Location

To

Location Start Duration End Train ID

GQ0107C 4471 LEG DALMUIR ARCHRAT 6:05 1:02 7:07 1Y11LB

GQ0107C 4472 LEG ARCHRAT ARDLUI 7:08 0:14 7:22 1Y11LB

GQ0107C 4473 LEG ARDLUI CRNLRCH 7:22 0:20 7:42 1Y11LB

GQ0107C 4474 LEG CRNLRCH BOORCHY 7:45 0:28 8:13 1Y11LB

FW0951C 4475 LEG BOORCHY RANNOCH 8:15 0:25 8:40 1Y11LB

FW0951C 4293 LEG RANNOCH FRTWLM 8:45 1:09 9:54 1Y11LB

Train graph

Methodology

• There are quite a few methods that may be applied in our project, we will take the first weeks to study the subject a bit and for choosing one algorithm to implement.

• evaluate the results: For this we will use a set of problem: – some very simple, so that we can verify that the correct solution is found– some more complex, so that we may not even be able to determine the optimal

solution but we can compare between different solutions. – Hopefully we will be provided with problem instances that also have a human-

created solution, so that we can determine whether our solution is as good as the human one.

Milestones

effort Due DateDecide what algorithm to use to solve the problem 2 weeks April 12Implement the algorithm 2 weeks April 26

Examine the algorithm on a small problem, check solution validity of quality.Data structure and data storage (import a real life large scale problem).

3 weeks May 10

Examine the algorithm on a large scale problem. 3 weeks May 31

Delivery, conclusions and project summary 1 week June 7

Development Environment

• Probably python or c++

Risks

• The selected algorithm may not be able to provide a good solution.

• The selected algorithm will have poor performance when solving a large scale problem.

Deliverables

• A working algorithm.• A list of valid and optimal diagrams.