Reducing Equipment Downtime and Maintenance Costs with a Predictive Maintenance Platform
The client is a leading global provider of intelligent material handling systems. Its customers include some of the top logistics, manufacturing, and e-commerce companies in the world. The company is driven by technological innovation and committed to providing its customers with best-in-class solutions.
A key line of business for the client is designing, building, and installing conveyor systems for their customers’ distribution facilities. The company places a strong emphasis on monitoring the health of its conveyor systems, maximizing their reliability and operational effectiveness and minimizing breakdowns and maintenance spending.
Even minor breakdowns can halt operations, resulting in lost revenues that can quickly escalate to over $1 million per hour for some customers. Given the cost of system downtime, interrupting operations to perform preventive maintenance that might not be needed or waiting for a part to break down before conducting repairs were not acceptable options.
DataArt was hired to design, develop, and deploy a predictive maintenance platform to reduce downtime and maintenance costs and maximize the operational effectiveness of the client’s conveyor systems.
Meeting the Challenge
DataArt team set out to build a cloud-based predictive maintenance platform capable of forecasting component degradation so that needed maintenance could be performed prior to equipment failure. The team consisted of a solution architect/machine learning expert, a data analyst, a back-end developer, a front-end developer, and four IoT developers.
One particularly difficult challenge was that conveyor systems were being installed in facilities with little or no connectivity and limited opportunity for wired connections. The team came up with a wireless mobile connectivity solution to route the data via gateways to the cloud and from there – to cloud storage. To ensure reliability, the team built mechanisms for gateway switching if one gateway were to fail.
Another challenge was that the team had to go beyond its usual scope of software work and tackle hardware development. The team installed sensors on all degradable conveyor parts, including motors, gearboxes, and bearings, to measure temperature, vibration, conveyor speed, power consumption, air flow, pressure, and other important variables.
In addition to capturing the real-time sensor data, the team gathered historical data, hardware-specific demographic information, weather and geographical data, inspection results, technical manuals, and maintenance reports.
The team also identified and selected the most suitable algorithms for training the machine learning models to predict the failure of each component part with the highest possible accuracy. Several machine learning models were developed for each component part, with each model responsible for monitoring a particular variable (temperature, vibration, power consumption, etc.). Each model was also trained to identify the range of “normal” behavior, determine whether any variances from the normal range are significant enough to warrant alerts, and determine their appropriate level of urgency.
This machine learning process is outlined in Figure 1.
Each model is based on the unique characteristics of each component part and its surrounding conditions. The predictive maintenance system for one conveyor is comprised of more than 220 conveyer/facility-specific machine learning models.
Even an identical conveyor set up in a different facility requires different models to account for different conditions in the external environment. The serverless, cloud-based solution developed by DataArt provides continuous monitoring of critical equipment, a real-time view of facility asset health, tools for data storage, analysis, and visualization, and a powerful alerting feature. Alerts for units at risk of failing are colored yellow or red according to the severity and proximity of the potential threat and are generated in rea ltime so that the plant maintenance staff can address equipment issues before a failure occurs.
The client and the facility’s maintenance staff interact with the system through a web-based dashboard.
The solution has the following characteristics:
cloud-scale data store
event data stream
no fixed costs and low variable costs
Security and governance
cognito and policies
Automatic disaster recovery
The project was completed in one year with four fully operational facilities in North America.
DataArt’s serverless cloud-based solution, which combines smart sensor technology, advanced data analytics, AI models, and machine learning algorithms allows the client’s customers to protect their high-value assets from unplanned downtime, maximizing equipment utilization, operational efficiency, and overall equipment effectiveness (OEE).
With remote monitoring of critical equipment and automatic alerts, repairs are performed only when and where needed. This reduces maintenance costs and allows for the most efficient allocation of maintenance resources.
When a conveyor system breaks down, not only can the loss of revenue quickly escalate to more than $1 million per hour, but the customer can suffer difficult-to-measure losses from customer dissatisfaction, reputational damage, and the weakening of its brand. By mitigating these risks, the predictive maintenance platform developed by DataArt enhances the client’s value proposition and promotes customer satisfaction.
Predictive maintenance also strengthens the client’s brand by helping to differentiate its systems and services from those of its competitors, very few of which offer predictive maintenance for conveyors.
Some of the benefits of a predictive maintenance system include:
The client can monetize the predictive maintenance solution by offering their customers additional analytics-driven services, including multiple dashboards, optimized maintenance schedules, or a technician dispatch service before parts need replacement.