Problem
Agbar, a leader in water management solutions, was confronted with significant challenges in maintaining the operational efficiency of water pumps in 200 deep water aquifers, ranging from 20 metres to 1 kilometre in depth.
The primary issues were:
- Connectivity Limitations: Network reliability was not assured at all sites.
- Bandwidth Constraints: Limited availability complicated large data transmission.
- Data Volume: Each pump produced about 70 GB of data daily at 20Hz, making centralised processing unfeasible.
The current infrastructure wasn’t suitable for traditional predictive maintenance solutions, which depend on sending data constantly to a central server for analysis.
Solution
Agbar partnered with OctaiPipe to implement an innovative FL-Ops approach on OnLogic Karbon 700SE-2 edge computing devices. OctaiPipe’s Federated Edge AI infrastructure made Edge AI solutions scalable, secure, and efficient, simplifying the machine learning model lifecycle.
The OnLogic Karbon 700SE-2 provided the computational power to process OctaiPipe’s predictive models and pump data locally. Data could then be analysed at the edge, on-site, with only key insights and model updates sent to the central system, leveraging OctaiPipe’s federated framework. This cut down bandwidth requirements and increased system resiliency to deliver an effective predictive maintenance solution.
OctaiPipe’s Impact
Prediction Maintenance: Predict failures before they happen, reducing downtime and improving water infrastructure resiliency.
Reduced Costs: Achieve 10x lower infrastructure costs compared to traditional centralised processing.
Enhanced Reliability: The system functions reliably in remote or poorly connected locations.
Operational Efficiency: Reduce water wastage and improve operational efficiency by 50x.
Agbar’s deployment of OctaiPipe in deep water aquifer pump maintenance brought several benefits. It predicted pump failures for timely repairs, improving infrastructure resilience and minimising downtime. Its dependence on local computation made it resilient to network issues, ensuring continuous operation even in remote areas. It also reduced water waste and increased the overall water management system efficiency.
About OctaiPipe
OctaiPipe is a platform for federated edge AI, focusing on trustworthy and scalable Edge AIoT development and management. Octaipipe’s Federated Edge AI solution for IIoT provides a secure, private, and efficient framework designed to be scalable, resilient, and cost-effective, making it a robust choice for decentralised critical Infrastructure applications.
Key Benefits
- Deployment of up to 500 devices
- Reduces learning cycles by 90%.
- Ensuring performance and resilience of smart device networks.
- Safeguarding sensitive IoT data at the edge, maintaining compliance and privacy standards.
- Decreased cloud dependency, reducing associated training, infrastructure, and data transfer costs.
- Capable of deploying up to 500 devices within 3 hours and cutting learning cycles by 90%.
- While minimising data transfer and compute expenses.
