Brainiall uses AI and AWS to help mining companies improve performance and sustainability

Using AI to improve industrial operations

Brainiall is a young company, launched with the goal of transforming industries by using scalable and replicable AI automation to help its customers improve productivity. This required the company to ingest and process large amounts of data to train machine learning models.

The ISV developed a cloud-hosted AI service that uses data provided through the internet of things (IoT) devices of its customers and other inputs to work like an autopilot. This integrates directly with supervisory control systems to collect sensor data, calculate optimal process parameters and make automatic adjustments to operations. Important parameters include mining flow, water flow, granulometry, input weight, output weight, the speed and power of each mill, temperature and humidity.

Brainiall’s AI automation software currently helps mining companies increase throughput in the grinding process by 5-10%. Grinding is the process where the large chunks of rock extracted from the mine are broken up to extract minerals. This increased throughput can increase revenues by millions of dollars and can cut energy use by about 3%, saving costs and reducing environmental effects.

Optimising product and planning for the future

Brainiall was looking for a partner to help develop and accelerate its SaaS transformation. Developing market-leading products and services, the ISV wanted to grow its business within the industry and beyond by tapping into new markets and broadening its customer base globally. At the same time, it also wanted to improve the efficiency of its AI software to be able to scale it for additional demand and support further growth.

SoftwareOne NextGen ISV Modernisation Programme with AWS

To achieve its transformation goals, Brainiall chose to participate in the SoftwareOne NextGen ISV Modernisation Programme with AWS, which is designed to accelerate and de-risk the ISV business transformation journey with guided product modernisation incentives and services.

The program enabled Brainiall to work with SoftwareOne to implement a new go-to-market (GTM) strategy to extend its AI-SaaS offerings to new customers through the AWS Marketplace. It also helped the company to migrate all of Brainiall’s operations to Amazon Web Services (AWS).

We chose to work with SoftwareOne and move to AWS because we believe the partnership will bring us significant technological and business benefits. By leveraging AWS' technologies together with SoftwareOne's expertise, we have improved the operation of our SaaS model. The SoftwareOne NextGen ISV Modernisation Programme with AWS has helped us reach new customers around the world with our AI services.

– Fabio Suizu, CEO, Brainiall

Brainiall had been using public cloud technology for a long time, but was facing difficulties. To ensure the best possible service for its growing customer base, Brainiall needed the latest technology to provide scalability, security and the ability to collect, process and analyse data in near real-time. It wanted more granularity in cloud services, with a focus on performance, flexibility, reliability, lean architecture and maturity in applying AI and machine learning. Looking at the cloud services in use, Brainiall also wanted to save cloud costs.

A multi-step approach to success

1. Envision Workshop

An envisioning workshop was conducted to conceptualise the needs and focus the business vision, transformation and innovation strategies, resulting in the following objectives that Brainiall wanted to achieve:

• Near real-time data processing and visualisation – dashboards to be updated every second
• Using ML and AI models trained on the data collected from IoT devices to optimise processes
• A leaner architecture, consolidating IT services and lowering management overheads
• An increase in performance and efficiencies
• Cloud cost optimisation: up to 20% cost savings on cloud consumption costs per month
• Easy and fast deployment of client environments
• Scalability for a growing customer base

2. AWS Assessment

To optimise Brainiall's AI SaaS solution for its mining customers, SoftwareOne carried out an assessment. It gathered the necessary information through interviews with Brainiall's IT team. The assessment aimed to evaluate the technical requirements for developing and operating a new and leaner architecture solution in the AWS cloud.

For the collection, analysis and output of IoT data, the objective was to evaluate how data lake, machine learning and data visualisation could best be implemented into the new architecture solution and which AWS services were best suited for this. As a result of the assessment, the main goal was to enable Brainiall’s SaaS solution to visualise data with a near real-time dashboard comprising eight variables with an update every second.

3. AWS Mobilise

In the past, Brainiall had a solution that was partially implemented in a public cloud. To unify information from two data sources and visualise the data, the company used another provider. Following the AWS Well-Architected Framework, SoftwareOne ultimately designed and developed a leaner and more cost-effective infrastructure based on the AWS SaaS model, using various AWS technologies to bring more performance to Brainiall’s IoT mining solution. After creating an MVP and test environment, SoftwareOne began using real data for processing and all the models defined for AI and ML. In addition, SoftwareOne deployed an AWS Infrastructure as Code (IaC) service using Terraform to automatically provision custom client environments.

Our AI services are relatively new to the market, but AWS and SoftwareOne were able to provide us with the technological capabilities to achieve our goals easily and quickly. Previously, we had several specific systems in place but we lacked control over them. By deploying an environment on AWS thanks to SoftwareOne, we now have a holistic solution. We can control the whole process and customise the cloud environment ourselves. This not only saves us development time, but has taken us to a whole new level of performance and flexibility. And we have been able to significantly reduce our cloud costs.

The following AWS services and technologies have been used:

IoT Core

AWS IoT Core provides the cloud services that connect IoT devices to other devices and to AWS cloud services. SoftwareOne used IoT Core route MTTQ messages, ensuring that all incoming messages can be integrated with other AWS services, such as Kinesis, which forwards message information for data storage in Amazon S3.

Amazon Kinesis Data Firehose

Kinesis Data Firehose is a streaming ETL solution which captures, transforms and loads data streams on Amazon S3, enabling near real-time analysis with BI tools and dashboards. It enables extremely fast and voluminous queries of data through sources such as Athena and ML/AI. As a fully managed service that scales automatically to keep up with the data transfer rate, it reduced administration for Brainiall’s IT team.

AWS Glue

AWS Glue is a serverless data integration service that makes it easy to discover, prepare, move, and integrate data from multiple sources for analytics, ML, and application development. SoftwareOne used the AWS Glue Data Catalog to create the data tables that were needed to search and query the data stored in AWS S3 using Amazon Athena and the Athena Connector.

Amazon Managed Grafana

Amazon Managed Grafana provides advanced data visualisation. Accessing AWS IoT SiteWise, it has been used to display near real-time dashboards, facilitating the development of new views and providing easy and secure access to different users and teams across Brainiall and its customers in different geographies.

Amazon SageMaker

Amazon SageMaker is a fully managed service for preparing data and building, training and deploying ML models. It has been used to illustrate the operation of ML in conjunction with other AWS components, focusing on how to extract and read data from sensors that will be available in Amazon S3.

4. Migration

This phase will see the migration of all production data to AWS and is the upcoming stage of close collaboration between SoftwareOne and Brainiall. The goal is to ingest over 100 IoT mining variables into AWS, and to process, parse and deliver to multiple services to provide the near real-time dashboard with a programmatic ML algorithm in Amazon SageMaker to automate the entire process.

The true power of our AI services lies in their ability to dynamically control and automate processes, optimising performance and efficiency. With up to an 8% increase in productivity, we consistently deliver exceptional value to our customers. The integration of cutting-edge AWS technologies further paves the way for our future success. SoftwareOne's extensive expertise in Cloud Services and Data and AI Services has been instrumental in our journey, and we are looking forward to continuing our successful collaboration.

– Fabio Suizu, CEO, Brainiall