Trucking Optimization
From UG
Contents |
Core Business Requirement
It is required to optimize total cost of domestic trucking business within one country (focus on USA now) for one company's supply chain for given time frame (let's say one year).
It is assumed that there is a constant flow of orders: O1, O2, ... , Oj, ...
Each order is about moving specific amount commodities (cargo) characterized by set of parameters: {Weight, Pieces (Num of palets + loose), Volume, Available Date, Due Date}
Commodities are to be moved together from Pick Up location to Delivery location.
For given Client Company there is a limited set of Up locations (origin doors) {od1, od2, ... , odN}
For given Client Company there is a limited set of Delivery locations (destination doors) {dd1, dd2, ... , ddM }
There are limited number of carriers (trucking companies) {tc1, ..., tcK} that can service country.
Each trucking company has its own tariff so we are dealing with {tct1, ..., tctK} - set of tariffs.
...
Suggested solutions
Number of things can be done to reduce total cost. For example:
- use economy of scale to move less frequently in as large truck as possible as full as possible
- use milk runs
- use hub and spoke
- etc (see more here: Business_Process_and_Use_Cases_for_NATP#Transportation_options)
- find carrier that charge less for selected trip
Phased approach
We follow phased approach which means that in every Sprint we add new optimization methods or improve existing ones.
Phase A
This phase was released to production on May 1.
In this first phase we added version 1 of Trucking_Optimization_Report. It is designed for Business_Process_and_Use_Cases_for_NATP#One_destination_and_milk_run_at_origin method.
Read Trucking_Optimization_Report article for more details.
Phase B
Number of features were suggested and partially designed - see below.
Trucking Optimization Report Version 2
In this phase it was suggested to add version 2 of Trucking_Optimization_Report. It is based on the concept of #Zones.
Main difference is to improve precision in defining "pick up zone" for milk run by defining zone not as a state but as an any number of pick up locations that are "not too far from each other" on the map.
See Trucking_Optimization_Report#Version_2_that_is_based_on_Zones
Zones
See attempt at designing Zone object here: Zoning.
Alerts
This would complement (or replace) user's task to run Trucking_Optimization_Report periodically or receive/review from Scheduler.
The idea is that every time CT is added - the system would "generate trucking optimization reports on a background". For every delivery location it would calculate total weight, pallets accumulated and available in each zone. If for any pick up zone "Ready to move conditions" is satisfied then system would alert operator by e-mail.
Ready to move conditions
There could be different options here but main idea is to define when we have enough goods to fill FULLY the truck or trucks.
Simplest example: total number of pallets is equal or more that the max number pallets that fit into smallest truck.
Ready to move conditions Admin
We need Admin for this conditions
Dashboard
Dashboard is simply a page where we summarize important current info, post alerts, etc.
This should be located on CT2 > Home (below top menu bar) - now this page is blank.
For this phase all we need is to post all alrts - see #Alerts
Phase C
Trips
We probably need this new business object in the system to support new features. It would describe one trucking trip from door to door with any number of pick ups and drop offs on the way.
To support this we need to add new fields to CT to be able to record any number of stops (with date/time, stop address, what commodities were picked/dropped, etc)
Visualization
- Google Maps - display moves, location on the map.
- Maybe Charts - TBD
Etc
Add "delivery zones" to support milk run at delivery.
Support all options from here: Business_Process_and_Use_Cases_for_NATP#Transportation_options
Use "rates module" (currently in development) to calculate rates for each carrier automatically, suggest cheapest to operator and display all if required.
Phase D
More automation.
New strategies.
Extensive use of math algorithms.
Find and apply best solutions found in industry and academia.
Preliminary Design
Preliminary Design Phase A
This phase is in production.
Preliminary Design Phase B
....
Preliminary Design Phase C
....
