Acerta uses machine learning to detect manufacturing defects in auto parts
Canada-based startup Acerta helps automobile manufacturers use data to learn about faults in manufacturing.
At a glance
Founders: Greta Cutulenco, Jean-Christophe Petkovich, Sebastian Fischmeister
Year it was founded: 2015
Where it is based: Waterloo
Funding raised: $2 million
The co-founders of Acerta worked together at the Embedded Systems Lab at the University of Waterloo in Ontario, Canada. Greta Cutulenco was a graduate student, Jean-Christophe Petkovich was working on his PhD, and Sebastian Fischmeister was the Professor in charge of the lab. They were doing research into fault detection and data visualisation, when several companies from the automotive industry approached the group asking them to apply their technology to their collected data.
Soon enough, the team realised that the technology they developed can provide significant value to the automotive and manufacturing industry, and decided to start Acerta.
Acerta offers a Software-as-a-Service (SaaS) platform that uses machine learning to provide real-time malfunction detection and failure prediction. Acerta’s proprietary cutting-edge technology allows clients to access the full potential of their collected data.
The platform learns the normal behaviour of the tested system and the complex correlations between data streams, and automatically detects anomalies in real time. It assists engineers with root cause analysis and provides them with actionable insight, thus saving millions of dollars in warranty claims, manufacturing down time, engineering resources, brand value, and legal expenses. Acerta’s clients include Fiat Chrysler Automobiles, Daimler and Volkswagen.
Greta Cutulenco, Co-founder and CEO says, "Vehicles are becoming more complex - both mechanically and electronically. This rise in complexity creates a strain on testing processes, which gives rise to quality issues, and eventually a spike in warranty claims and recalls. Acerta offers a SaaS platform that uses machine learning to provide real-time malfunction detection and failure prediction. The platform learns the normal behaviour of the tested system and the complex correlations between data streams, and automatically detects anomalies in real-time. This enables manufacturers to utilise all of the data they collect to produce accurate insight into their system quality."
Acerta's core technology is a machine learning algorithm for defect detection and failure prediction using machine data. The technology is used by product design and development engineers for pre-production testing, by manufacturers for end-of-line and quality testing, and by asset owners and fleet managers for predictive maintenance.
The company incorporated in October 2015 and the first product was delivered shortly after. Since then, the startup has worked on various successful projects and are on the way to full deployment with several manufacturers. Acerta has raised a seed round of $2 million in July 2017 from a group of investors including Spectrum 28, Garage technology Ventures, Trifecta Capital, Plug and Play and OMERS ventures.
They are currently in the process of expanding their reach in the US, Canada, and parts of Europe. “India is fertile ground for us with its massive manufacturing industry and high level of innovation, and we are considering hiring a regional sales executive out of India after we raise Series-A funding,” adds Greta on their plans for India.
Market intelligence provider Trendforce predicts the global market for smart manufacturing (industrial IoT and artificial intelligence) solutions will grow at a CAGR of 12.5 percent from 2017 to 2020 and surpass $320 billion by 2020.