1	Presented by: Kevin Callen WasteBid.com, Inc. Presented for: WasteExpo 2003 Using GIS Routing Software to Improve Collection Efficiencies
2	About WasteBid.com Dot com survivor Evolved from e-commerce website to software developer Applications include: FleetRoute - route optimization software SmartEngine – online marketing and waste audit software MatTrack – full cost accounting software Application/Website development and web-based marketing
3	In This Presentation Goal: Develop industry understanding of route optimization technology Outline What is the Technology What is not Route Optimization Technology Benefits What to Look For in a Routing Software What to Beware Of Why Bother
4	What is Route Optimization Software? Modeling application Hopefully better than weather models!!! Balances route times, service days and districts Reduces labor/overtime costs Optimizes travel paths through streets and to disposal facilities Decreases fuel/tire costs and mileage
5	Balance Service Days and Districts
6	Nashville Collection Days - Before
7	Nashville Collection Days - After
8	Create Routes Maximizing Each Vehicle’s Productivity
9	Optimize Travel Paths
10	What to Expect 10-15% reduction in operational expenses Reduced number of trucks/crew 25% or more savings if coupled with other factors: new disposal facility or collection system Difficulties to accomplish this Street and customer data must be complete Travel path algorithm must be very sophisticated to deal with the many variables in reality Finalizing draft routes by drivers and supervisors and easily getting their changes back into the routing software
11	Typical Results – Mill Valley Refuse Service
12	Other Benefits: Improves Operations Integrates routing with customer service and billing Improves quality control and management Maintains customer service times and information Increases accountability of crews Standardizes procedures for routing, training and dispatching “Key Man” insurance policy Upgrades GIS street data and leads local government initiatives in GIS for other agencies to use the new data Emergency response School buses, meter reading, etc.
13	What technology does not effectively optimize routes? Route management, dispatching, billing software Basic area routing – counting houses in GIS Route paths using MapPoint GIS Should use high-end GIS of ArcView, MapInfo, etc. for complex modeling GPS vehicle location systems Data can be used for route optimization
14	What To Look For in a Routing Software Sophistication Algorithm that mimics reality Automation Flexibility Adapt to real world problems
15	Understand Algorithm Types High density residential collection routing Must service nearly every street Biggest potential for saving money This presentation’s focus Commercial collection routing Point2point routing Easier problem Dynamic/Real-time Routing Most products only do one or the other
16	High Density Residential Routing Application Areas Public Works & Utilities Waste/recycling collection Snow plow, street sweeper, bug spraying Meter reading Postal Deliveries Political Redistricting Other Analytical Uses Collection cost analysis Districting Facility site evaluation
17	Point-to-Point Routing
18	Look for Attributes of a Sophisticated Algorithm Customers on side of street Single and double pass in same route Time-specific streets/customers Avoid school or rush hour traffic U-turns Trucks can’t u-turn on certain roads
19	More Attributes of a Sophisticated Algorithm Individual customer set-out weights and service times Individual truck sizes/capacities Disposal trips: closest point vs. when full Multiple disposal facilities Go to closest or cheapest?
20	What Else to Look For Maximum size of problem solvable by the system Number of Stops Number of Vehicles Number of Depots Number of Disposal Facilities Size of Street Network
21	Nature of Routing Problem Sizes Too small, then time and cost does not easily justify effort Typically at least 4-6 trucks, depending on reasons for routing Reverse Economies of Scale Larger the problem, the more difficult to solve and the more costly on a per route basis If data clean-up is an issue
22	Simplification Through Automation Route Generation Automation Automated Map Production Billing/Customer Service Data Importing and Automated Address Matching Automated Route Updating
23	Why Automation? Systems Without Automated Processes Time consuming to update routes Time consuming to generate/print maps Reduces time and cost of updating routes and rerouting More easily integrate new customers; facilitate expansion
24	Flexibility Through Object Oriented Programming
25	Flexibility to Solve Real-world Problems Neighborhood Routes - uses major streets to create routes that maintain neighborhood boundaries Time Sensitive Routes - allows you to merge sub-routes together to handle priority streets and other unique situations
26	Uses for Routing Considering Time Sensitive Collections Schools Major streets before traffic gets too heavy Streets that have parking restrictions at certain time periods to allow easier access to garbage Hospitals and other “quiet zones” Customers with time-specific collection requirements
27	A Priority Route Around a School
28	Flexibility: Standardized and Custom Outputs Color driver and management maps Street-by-street and minute-by-minute driving directions Statistical reports for each route Profitability reports for each route Customer change of day lists for notification Ad-hoc mapping and analysis function Analyze customer pricing and targeted marketing
29	Beware: Purchasing Evaluation IT staff may have responsibility to make purchasing decision Evaluation must weigh the benefits of cost savings Are shortcomings bad enough that the technology will not be effectively implemented to realize those savings
30	Beware: Cleaning Up Existing Data Big downfall of routing projects Second only to poor unrealistic routing algorithms that make poor/unusable routes Street data purchase versus clean up existing street data (if public sector GIS) What brand of map is purchased? NavTech, Tele Atlas are best (GDT is OK) If Census Tiger – routing not sophisticated for high density residential Good area to use assistance of consultants that are very knowledgeable of data clean-up processes Use scripts to automate tasks Software that aids in identifying and solving data problems
31	Beware: Vendors Vendors that say they can do everything Being sold a product that can’t do your type of routing, e.g., high density residential versus commercial point2point routing Vendors that say routing is simple and easy Be prepared for demands on technical staff Getting sucked into thinking its easy and then going over budget to complete the project or just not completing it at all
32	Beware: How Long Will It Take? Months, Years, Never (failure) Best case: 2-6 months Depends upon: Data quality Quality of algorithm/software functionality GIS expertise of staff Assistance from vendor Experience of vendor Commitment/allocation of resources of drivers, supervisors and managers
33	Why Bother? Save Serious Money in Collection, which account for up to 2/3 of operations 10-15% savings with unbalanced routes Up to 25% with new disposal facility Improves operations and customer service New Competitive Advantage WM & BFI testing nationwide roll-out
34	Thank you for your time Telephone: 866-WASTEBID Email: kcallen@wastebid.com Web: www.wastebid.com