Holand Technology - Intellectual delivery system for transportation industry #HackingSAS
| Team Name | Holand Technology |
| Track | Data for Good |
| Use Case | Intellectual delivery system for transportation industry |
| Technology | SAS Viya, Python |
| Region | AP |
| Team lead | Chen Jiasheng |
| Team members | @yzm @tanghc @feijizhadan |
Who we are
Founded in 2005, Holand Information Technology Co., Ltd mainly engages in data mining, machine learning and artificial intelligence. Its intellectual products have empowered a wide range of industries, including government, finance, healthcare, education, manufacturing, retail, operators, and more. At the same time, its technology community has been training big data and artificial intelligence for enterprises for more than a decade.
Introduction
With the rapid growth of goods movement demands in retail and consumer packaged goods, transport companies are under intense pressure with their limited resource. The number and species of goods vary greatly at different places in an area according to factors such as economic development and population size.
Many transport companies now use a set of human-dependent distribution delivery program. Firstly, it divides more than tens of thousands of sales points into some areas based on geographic location. Then one car is responsible for the orders of all points of sale delivery in an area every day, with the routes generated by common navigation software. Although this program is convenient and usually fits the demand, it is prone to problems when the demand for goods increases. For example, over the festivals, the demand for gifts is large, and the economically developed regions will need more goods, while factory areas may need less because of the rapid decline of the population.
What we want to do is to build an intellectual delivery system. With this efficient system, the transport companies can effectively use existing vehicle resources to complete daily distribution tasks, and can carry out new distribution services, distribution of their online shopping mall purchase orders, such as wine and snacks and so on. This can greatly save resources and achieve greater benefits.
Not only the retail goods distribution, but other industries also need such an efficient system, such as the car-hailing service and in-city delivery services. This system helps improve distribution efficiency and reduce exhaust emissions to be more environmentally friendly.
Technique
Specified a few points, how to find out the fastest route to pass through all of them is a problem that has plagued the mathematical academic community for nearly a hundred years, has not yet been solved, called NP complete problem. Our method also does not find the optimal solution with proof, because we can not traverse all possible routes, but it is much better than the current method.
What we used is genetic algorithm, a method of searching for optimal solutions by simulating natural evolution. With some set optimization strategies, we can dynamically balance the problem according to demand and resource ability, and quickly improve the reaction speed.
Result
Here's a sample diagram in attachment.
Look at the gif chart, the yellow dots denote the points of sale that needs goods for the day, the big red dot denotes the distribution center, and it also shows the delivery orders and routes which are marked green designed by our system.
