We are sAInce.io
Who we are.
Each of us has more than 20 years of experience in automation within Manufacturing.
Over the years we gained experience in IIoT, Micro Service development, DevOps, Big Data and Machine Learning.
We build further on our Network of peers to bring you the best in Smart Manufacturing.
Currently we are looking for seasoned peer experts and junior enthusiasts who blend well with our culture to extend our team of high-end consultants.
See our Job page for more information.
We build intelligent edge devices around plain sensors, PLC’s, … and equipment, tailored to your business needs.
Over the air monitoring, configuration, … automatically deployed from the cloud or your own infrastructure. We make any device or equipment smart.
Open up the world of Machine Learning to your devices.
In most cases it’s not enough to make devices smart but you’ll want to incorporate business logic that is tailored to your manufacturing processes.
We build micro services so you can deal with this logic without compromising the smart edge device.
Think about real-time cockpit applications for alerting and responding to process events, anomalies and predictive maintenance.
Act immediately on real-time data streams.
Visualize actual and predicted trends and apply ML models in real-time.
We can help you in building these micro services and applying best practices for modularity, reuse and integration with external applications.
We are specialized in AWS Serverless applications
Provide access to all your users on integrated data in the cloud. We build applications in Angular with backends in Node-Js and/or Python through lambda functions.
We develop the AWS architecture and can integrate all your requirements with the full AWS services stack.
We assist you towards a full serverless architecture.
We have extended knowledge in the field of application integration and data integration.
We augment historians so they can be enriched with ERP, MES, LIMS, WMS etc data for full insight in the equipment data.
We develop event driven applications that provide real-time ERP, MES etc data to other applications or services so systems remain loosely coupled but enriched with each others data.
Before you can actually make devices or applications smart you need to train and test large datasets. Enters the world of big data.
We build data pipelines to inject your data into Cloud native platforms (Public, Private or On-Premise) and transform them into fast and usable formats for Data Science and Business Intelligence consumption.
Whether it’s about historical backfill operations or real-time data ingestion, we automate your data lake so that your Data Scientists don’t have to worry about the data and can focus on Data Exploration and ML model training and testing.
We assist you in the selection of the appropriate storage medium (NoSQL), and set up data architectures for optimal loading and retrieval of data.
Deep Learning (AI)
We develop deep learning solutions. Think about object detection, anomaly detection and preventive maintenance.
We use Convolutional Neural Networks, Generative Adversarial Networks, Auto-Encoders and Long Short Term Memory networks.
Models are trained in the cloud on fast distributed GPU-accellerated machines.
We specialize in the development of Deep Learning models with Pytorch and Keras and deployment to Edge Devices.
How we work.
IntakeIn general, we start with one or more workshops to gather as much info as possible regarding the business case.
The focus is on formulation of clear business goals, business questions, timeline, data sources, owners and availability of resources.
The final result is a definition of success agreed by all stakeholders.
This phase takes a couple of days but could be spread over 1 to 2 weeks.
PoCThe next step is to build a Proof of Concept that proves that there is a potential solution to the problem at hand.
Depending on the complexity of the business case, it's possible that this phase first requires extra tasks like data exploration, more in-depth interviews with key stakeholders, investigation of existing technology stacks and/or the available devops environment, etc.
This phase takes approximately 1 - 2 months.
Agile DeliveryOnce the result of the PoC receives a GO, we enter in a delivery stage that follows an agile approach. Practical but not over-complicated. You decide what functionality is key and what needs to be developed first.
A backlog is defined and delivery is done in sprints, usually of 2 to 3 weeks. Each sprint cycle is finished with a sprint review and decisions for the next sprints.
This phase can take multiple months.
You, the customer, decide on the scope of the project and what will be delivered each sprint. This gives you the opportunity to remain in budget while still focusing on the key business goals of the project.