Startup develops automated materials handling system29 de outubro de 2019
By Eduardo Geraque | FAPESP Innovative R&D – Seven years ago, two friends who were undergraduates at the University of São Paulo’s Engineering School (POLI-USP) in Brazil and shared a passion for robotics founded Automni to develop innovative solutions for the automation of materials handling.
The startup’s flagship product is Rhino, an autonomous pallet truck and trans-stacker controlled by artificial intelligence and used by such large companies as Unilever, Mercado Libre and Dell in manufacturing plants and distribution centers. “A system coupled to the vehicle receives orders electronically, identifies the desired product based on the address provided – an aisle or other location in the distribution center – and delivers it to the correct destination,” says André Abrami, one of the firm’s founders. “Our current priority now is to raise more funds from investors in order to ramp up production.”
The firm plans to extend the functionality of the technology used in Rhino by adding automatic inventory control. The project is supported by FAPESP via its Innovative Research in Small Business program (PIPE).
The new inventory control system is still being developed but has already been branded Argos. The idea is to couple a mast to Rhino so that the robot can maintain a complete inventory of the goods stored in a warehouse or distribution center. “If it takes a couple of days to inventory storage now, our intent with the new system is to complete the inventory during the staff’s lunch break,” Abrami says.
To achieve this goal, the firm will install a set of cameras on a retractable telescopic mast to read the barcodes on the labels of stacked crates. In a distribution center with seven-story racks, for example, two cameras per story will be needed for each side of the aisle, making a total of 14 cameras on the mast, according to Abrami. In distribution centers with taller racks, the mast will also have to be taller, and more cameras will be required.
The main technological challenge to be surmounted in the months ahead will be calibration of the optical instruments to assure accurate image capture in any lighting conditions. “Accuracy of barcode capture will also have to be combined with the right speed of the robot on the ground,” Abrami notes.
Manual barcode reading
Managers of distribution centers must take inventory at regular intervals for legal reasons as well as to guarantee proper control. In Brazil, according to Automni’s data, most do so by manually capturing barcodes. Depending on the volume of goods held in a warehouse, manual data capture is time consuming and can lead to accidents since the employees who perform this task often have to work at height.
Furthermore, Abrami adds, the use of a system that does not require human presence assures accuracy and avoids data processing errors. “We will develop the masts and image capture system, all of which will be operated by the software that already runs Rhino,” he says. Images will be classified by neural networks.
The fact that the founders of Automni have already found a niche market for the application of robotics has won them international recognition. The firm’s presentation was awarded first place in the Brazilian stage of the third edition of the Startup World Cup, held in São Paulo in October 2019 during the 7th Brazil-Germany Innovation Conference. In Brazil, ten startups participated in the event, which was organized by Pegasus Tech Ventures in partnership with Invest SP, the São Paulo State investment and competitiveness promotion agency. This round of the Startup World Cup took place in 40 cities around the planet.
By winning the Brazilian stage, Automni advanced to the grand finale, to be held in Silicon Valley in May 2020, and will have the opportunity to compete for an investment prize of USD 1 million.