Geo-Spatially Optimized Order Scheduling
Geo-Spatially Optimized Order Scheduling How to Save a Million Miles
TXU Operational Highlights ! Manages over 100 work types from CIS, Maximo ! ! ! !
~ 800 Techs (400 Gas, 400 Electric) ~ 20 Dispatchers (50:1 actual dispatcher to technician) ~ 100 Field Supervisors ~ 150 Managers
! Work Orders ! ~ 17,000 to 25,000 per day (35,000 potential) ! ~ 3.5 to 4 Million orders per year ! 250-500,000 transactions per day with CIS
! Gas and Electric managed by same system ! Common system administration ! Common wireless infrastructure ! Common server with multiple hosts and wireless connections
Challenges of Automated Assignments ! Large Territories ! Cities (dense areas) ! Rural (sparse areas) ! Multiple Work Types ! Multiple Skill Types ! Natural Boundaries
! Political Boundaries (unions, service centers) ! Workload Utilization ! Workload Balancing ! Emergency Response ! Priority and Appointment work
The Objective ! ! ! ! ! !
Maximize utilization of field technicians/crews Work the highest priorities on-time Minimize travel time Minimize administrative effort Minimize computation time (efficient algorithms) Monitor results, manage exceptions
The Solution ! Geo-code orders as needed for lat-long coordinates ! Balance the day’s workload ! Geographically optimize order assignments ! Assign FSR to concentric clusters of work ! Minimize FSR ‘cross-over’
! Apply street level routing sequences for best travel time ! Meet appointment requests ! Ensure top priority work is completed
! Dispatcher maps – ‘see’ your fleet and workload ! Supervisor maps – know where your crews are! ! Technician mobile maps – minimize prep time and missed orders
TXU Gas Case Study ! Establish large areas equivalent to the max orders and techs a dispatcher can effectively manage (e.g. 50-60 techs/disp). ! Use Latitude/Longitude coordinates between technician’s start location and the order’s lat-long to calculate closest tech. Geo-code as needed. ! Algorithm ! Calculate amount of available work and workers – even distribution in this instance ! Consider techs ‘qualified’ for the work (must meet eligibility rules for skill, schedule, etc.) ! Use a “governor” to limit distance traveled (also helps with natural & political boundaries) ! Assign the work
Results ! Automatically builds non-overlapping concentric routes based on order location and order volume ! ‘Moves’ the technicians to the work – increases worker utilization ! Recommend / Suggest feature uses same algorithm. ! Improved windshield / travel time (0.6 miles / order) ! Reduced pre-shift adjustments and prep time over 90% 6 planners, 8 hrs each
2 planners, ½ hr each
! ~90% of orders are automatically routed ! discrepancies are typically bad lat-longs or no qualified tech
2-Step Approach to Optimized Travel Time ! Workload Distribution using Geo-Spatial optimization assigns orders to technicians based on location, skills, and availability to minimize tech crossover and maximize utilization ! Street Level Routing (SLR) sequences each technician’s assigned orders to minimize travel time and ensure appointments are met.
2 1
2 1
Processing Assignment Geo-Spatially Each Tech starts at pre-determined ‘home’ location
Map Area
T1
T1
T2 gest
on D-1l
T2
D-2l
ong est
T1 3
1
T1 T2 est long 1 T2 D
T2 2
T1
4
Processing Assignment Geo-Spatially Techs starting at ‘same’ location
T1 T2 T1
D T2 -2longest
T2 3
1
T1 est
long T1 D-1
T2
T1 2
T2
4
Case-Study Example: Tech ‘home’ Location
Case-Study Example 1
Case-Study Example 2
Street Level Routing Travel-Time Optimizing ! Geo-codes orders ! Determines the sequence of orders ! Determines ETA and ETC for each order ! Provides driving directions ! Enables dispatchers and technicians to view routes
4
5
1
3
2
Lessons Learned ! Use current GPS/last known position ONLY for emergency assignments to prevent crossover ‘creep’ ! Use map view of order assignments (dispatcher and /or supervisor) to quickly verify proper assignments ! Improve administration tools (e.g. map-based entry of tech lat-long vs. keyboard entry) ! Add Geo-Coder during order download when lat-long not present on order (otherwise all orders must have valid lat-long) ! Use map assisted reporting for assignment and optimization analysis
Processing Assignment Geo-Spatially Techs starting at ‘same’ location
T1 T2 T1
D T2 -2longest
T2 3
1
T1 est
long T1 D-1
T2
T1 2
T2
4
Dispatcher View of Orders and Field Crews/Techs
Street Level Routing for Mobile Technician
Mobile Mapping Features View Map
View Info
! Provides Work Order context from Advantex ! Manages Work Orders from the map
View SLR Route
Powered by ARC-GIS Find Object
View Directions
View Assets
Thank You
TXU Operational Highlights ! Manages over 100 work types from CIS, Maximo ! ! ! !
~ 800 Techs (400 Gas, 400 Electric) ~ 20 Dispatchers (50:1 actual dispatcher to technician) ~ 100 Field Supervisors ~ 150 Managers
! Work Orders ! ~ 17,000 to 25,000 per day (35,000 potential) ! ~ 3.5 to 4 Million orders per year ! 250-500,000 transactions per day with CIS
! Gas and Electric managed by same system ! Common system administration ! Common wireless infrastructure ! Common server with multiple hosts and wireless connections
Challenges of Automated Assignments ! Large Territories ! Cities (dense areas) ! Rural (sparse areas) ! Multiple Work Types ! Multiple Skill Types ! Natural Boundaries
! Political Boundaries (unions, service centers) ! Workload Utilization ! Workload Balancing ! Emergency Response ! Priority and Appointment work
The Objective ! ! ! ! ! !
Maximize utilization of field technicians/crews Work the highest priorities on-time Minimize travel time Minimize administrative effort Minimize computation time (efficient algorithms) Monitor results, manage exceptions
The Solution ! Geo-code orders as needed for lat-long coordinates ! Balance the day’s workload ! Geographically optimize order assignments ! Assign FSR to concentric clusters of work ! Minimize FSR ‘cross-over’
! Apply street level routing sequences for best travel time ! Meet appointment requests ! Ensure top priority work is completed
! Dispatcher maps – ‘see’ your fleet and workload ! Supervisor maps – know where your crews are! ! Technician mobile maps – minimize prep time and missed orders
TXU Gas Case Study ! Establish large areas equivalent to the max orders and techs a dispatcher can effectively manage (e.g. 50-60 techs/disp). ! Use Latitude/Longitude coordinates between technician’s start location and the order’s lat-long to calculate closest tech. Geo-code as needed. ! Algorithm ! Calculate amount of available work and workers – even distribution in this instance ! Consider techs ‘qualified’ for the work (must meet eligibility rules for skill, schedule, etc.) ! Use a “governor” to limit distance traveled (also helps with natural & political boundaries) ! Assign the work
Results ! Automatically builds non-overlapping concentric routes based on order location and order volume ! ‘Moves’ the technicians to the work – increases worker utilization ! Recommend / Suggest feature uses same algorithm. ! Improved windshield / travel time (0.6 miles / order) ! Reduced pre-shift adjustments and prep time over 90% 6 planners, 8 hrs each
2 planners, ½ hr each
! ~90% of orders are automatically routed ! discrepancies are typically bad lat-longs or no qualified tech
2-Step Approach to Optimized Travel Time ! Workload Distribution using Geo-Spatial optimization assigns orders to technicians based on location, skills, and availability to minimize tech crossover and maximize utilization ! Street Level Routing (SLR) sequences each technician’s assigned orders to minimize travel time and ensure appointments are met.
2 1
2 1
Processing Assignment Geo-Spatially Each Tech starts at pre-determined ‘home’ location
Map Area
T1
T1
T2 gest
on D-1l
T2
D-2l
ong est
T1 3
1
T1 T2 est long 1 T2 D
T2 2
T1
4
Processing Assignment Geo-Spatially Techs starting at ‘same’ location
T1 T2 T1
D T2 -2longest
T2 3
1
T1 est
long T1 D-1
T2
T1 2
T2
4
Case-Study Example: Tech ‘home’ Location
Case-Study Example 1
Case-Study Example 2
Street Level Routing Travel-Time Optimizing ! Geo-codes orders ! Determines the sequence of orders ! Determines ETA and ETC for each order ! Provides driving directions ! Enables dispatchers and technicians to view routes
4
5
1
3
2
Lessons Learned ! Use current GPS/last known position ONLY for emergency assignments to prevent crossover ‘creep’ ! Use map view of order assignments (dispatcher and /or supervisor) to quickly verify proper assignments ! Improve administration tools (e.g. map-based entry of tech lat-long vs. keyboard entry) ! Add Geo-Coder during order download when lat-long not present on order (otherwise all orders must have valid lat-long) ! Use map assisted reporting for assignment and optimization analysis
Processing Assignment Geo-Spatially Techs starting at ‘same’ location
T1 T2 T1
D T2 -2longest
T2 3
1
T1 est
long T1 D-1
T2
T1 2
T2
4
Dispatcher View of Orders and Field Crews/Techs
Street Level Routing for Mobile Technician
Mobile Mapping Features View Map
View Info
! Provides Work Order context from Advantex ! Manages Work Orders from the map
View SLR Route
Powered by ARC-GIS Find Object
View Directions
View Assets
Thank You
