State of Goiás Department of Economy uses AI to fight tax evasion in freight transport

The Brazilian government agency uses SAS Enterprise Guide and SAS Visual Analytics on SAS Viya for cross-referencing and visualizing data to generate auditing alerts

The Department of Economy of the State of Goiás has adopted SAS technology to combat tax evasion in freight transport. The initiative launched in April and, within one month, allowed the Brazilian government agency to increase fines by 54% compared to the same month last year.

The project aims to inspect goods in transit and uses monitoring and artificial intelligence technologies to cross-check tax data in real-time. Optical character recognition (OCR)-equipped cameras on antennas throughout the state’s highways scan vehicles' license plates.

The antennas track approximately 30,000 cargo trucks traversing the roadways daily. They transmit data from license plate photos to a central office, which cross-checks it with information on the goods’ invoices. The system alerts on any suspicious movement.

SAS® Enterprise Guide® supports the data cross-referencing to issue alerts. Meanwhile, SAS Visual Analytics running on SAS® Viya® provides data visualization to track alerts. SAS Viya is a cloud-native AI platform that provides customers with options and controls for analyzing large, complex data sets with the required scalability.

With SAS Enterprise Guide, business analysts can produce their own analyses and reports, freeing IT staff to focus on other strategic projects. About 14 auditors trained to use the solution. As a result, they can leverage the systems analysis and reporting capabilities and use the technology more successfully, according to Giovana Amorim Zanato, Tax Auditor of the State Revenue and Manager of Prospecting and Auditing for the Department of Economy of the State of Goiás.

The department also used SAS Enterprise Guide to create a logistic regression using advanced statistical methods to identify vehicles that could be reporting only part, not the entirety, of their cargo. With SAS, the department also developed a model to indicate which trucks are likely to be overweight so that auditors can be alerted.

With the use of SAS Visual Analytics on SAS Viya, the department can visualize data in real-time, streamlining a process that would previously update every 30 minutes. Moreover, the dashboards were not associated with each other before, meaning it was impossible, for example, to insert a filter in all dashboards at once. The dashboards are monitored at the Department of Economy’s State Operations Center, COE.

Benefits to the citizen

In its first month of operation, the system generated 300 alerts of possible tax inconsistencies per day with the help of SAS, identifying irregularities including forged invoices, missing electronic manifests mandated for freight transport vehicles and cargo exceeding the declared quantity. If a tax violation is found, the due tax is collected from the carrier, in addition to applicable penalties, curbing tax evasion and unfair competition.

“This project has greatly changed the department's operational efficiency, bringing benefits to society and increased transparency for the department,” said Zanato. “We combined SAS technology with other technologies to prevent tax evasion and created something innovative to serve as an example for other states' departments.”

Other state offices have approached SAS for solutions for similar projects. “Today, the public sector is one of SAS' most important areas of activity in Brazil. We have the opportunity to bring the most advanced technologies, such as artificial intelligence and machine learning, to the national market with fully replicable solutions,” said Marcelo Laranjeira, SAS’ Head of Government and Public Banks in Brazil.

He underscores that technologies such as artificial intelligence are used increasingly by managers in decision-making and improving public services. “These resources can make Brazil a country better equipped to face its public sphere challenges, creating a truly data-driven culture in the sector and benefiting citizens with efficient services,” said Laranjeira.

This is an English translation of a news release issued by SAS Brazil on Aug. 31, 2023. See original announcement.