Projects

Employees of any company interact with various types of software products during the working day (such as Mail services, ERP systems, industrial CRM). They keep records in 1C, generate reports in corporate portals, use EXCEL, look for information in browsers and search engines, etc.. About 50% of work performed by employees is a routine - repetitive work with a simple decision-making algorithm.

When using RPA, work will be done faster, a cost of production or services and an amount of errors will be reduced, common standards will be introduced in various departments and the level of motivation of employees will increase.

IТ-company MigЗЗ is a certified partner of ElectroNeek platform. We consider this direction very promising, and partnership with the team of Sergey Yudovsky is strategic for Mig33. 

I am Sergey Votyakov, mig33.ru expert, and I will tell you how we implemented RPA technology in logistic company.

✔ How it was?
The company generates more than 8000 product declarations, more than 2000 contracts are analyzed and checked annually. Managers collect and compile all documentation manually.
They turned to us with a request: to robotize the process of filling out a customs declaration.


What the process consists of? 

⦁ To download data from 1C
⦁ To process & to select the needed data
⦁ To fill out a customs declaration

Data can be stored in PDF and PNG format, this makes data transfer complicated, you have to enter data manually.
Of course, this takes a lot of time, and moreover, with the increase in orders, the probability of making a mistake increases too.

✔ How it became?
The task was to make software for the formation of the customs declaration and 1C "be friends".

The process should look like this:

⦁ To download data from 1C
⦁ To recognize images
⦁ To gather the necessary information from PDF and PNG formats
⦁ To insert data into a customs declaration
⦁ To check data and to send a declaration


✔ Steps

Step 1:
     To set a task, to discuss it with all of the process participants and to create a process map.
Step 2:
     To evaluate capabilities of process robotization.
Step 3:
     To develop a robot and to test it.
Step 4:
     To make a detailed test of the robot's abilities in non-standard situations, to improve it's prototype
Step 5:
     To start it in production mode, to develop new functionality according to customer's needs.

✔ Difficulties

When developing the system, we encountered such difficulties:

⦁ Recognition of PDF and PNG file formats
⦁ Lack of a single standard for supporting documents
⦁ Files' names were changed (Human Factor)

✔ Solutions of our team:

⦁ We used GhostScript, which helped us split PDF files into pictures, and then search and isolate the data we need. This allowed us not to increase the budget of the robot development.
(in the picture the robot finds the name of the company and enters it into Excel)


⦁ To solve the problem with the lack of a single standard, we proposed to separate the general types of orders and documents. Now this solution allows one to accurately find the necessary documents.

⦁ Now the robot determines the names of the necessary documents using regular expressions.
(In the picture, the robot sorted the documents and put them in a separate folder)


✔ Results