Follow these simple steps to get great results...
Amazon Monitron is a Predictive Maintenance (PdM) platform that sets a new bar for low cost and simplicity of operation without sacrificing the power of predicting asset failures. We cover Monitron's amazing benefits in an earlier blog post. In this post we cover the techniques to extract maximum performance in both a Pilot and deployment, and tips that make Monitron easier and more powerful. They are based on our direct experience with over 14 Pilots and POCs, multiple PdM platforms including with Monitron where we have helped our clients achieve an ROI of 60x profits/cost. This blog post is complementary to the Amazon Monitron Documentation located here...
WHY IT MATTERS: AI-based predictive maintenance is already having a significant impact on plant operations. It is imperative to manage your PdM platform on an on-going basis to ensure continuous high ROI. The more ROI you can demonstrate, the more budget you will secure for scaling the PdM at multiple sites.
THE FOCUS: We will be focusing on predictive maintenance for assets in industrial operations. We provide special tips and insights for Amazon Monitron. These steps can be adapted to virtually any AI-based predictive solution.
Monitron Pilot, PHASE 1 - Setup (14-30 days)
A checklist to reach optimum asset selection and sensor assignments to support the Pilot.
1. Build the team:
Maintenance engineering and technicians – they help with asset selection, sensor installation, alarm monitoring and making decisions when to release an asset for repair.
Data science team – Not required for Monitron. All the AI modeling is done automatically through internal services based on AWS SageMaker service. Its settings cannot be configured. Other platforms will require data science staff for alarm customization
IT Staff – Just one Certified Solutions Architect to set up the AWS account and turn on AWS Monitron system. DecisionIQ provided this service.
Partner for experienced coaching – Managing and AI-based Pilot is as much art as science. An AI experienced partner will help you set the right expectations, maximize ROI by finding the best opportunities and avoid pitfalls unique to AI-based alarms. This is the role that DecisionIQ plays in its many successful Pilots. AWS will often fund DecisionIQ coaching in the Pilot as a service to its accounts.
2. Select the Pilot site based on its maintenance culture:
[good] primarily a run to fail culture with limited work orders, technicians with anecdotal histories. Monitron is ideal for this culture because it provides an immediate step-up into condition-based monitoring and alarming with a dashboard that is easy to use and interpret.
[better] a mix of run to fail and preventative maintenance (PM) culture with work orders for some key assets, experienced technicians with anecdotal histories.
[best] a reliability-based culture with condition-based monitoring and PM practices, diagnostic support staff and disciplined work order system with two-year history. Monitron is a powerful tool to scale asset coverage for these PM teams, shifting them into heavy diagnostics roles.
3a. How Many Sites?
Why focus the Pilot on a single site?
To increase probably of capturing wins, concentrate budget and sensors at one site.
Concentrate training and reporting, start building a scalable PdM culture.
Why focus on multiple sites?
To get results based on different types of assets and maintenance cultures.
To engage other personel for training and reporting.
Once site(s) have been selected...
4. Gather asset maintenance data, 1-2 year histories specific to that site:
Work orders, all types including corrective, preventative and emergency.
Downtime report for key assets.
Asset hierarchy for all assets.
5. What sensor data history is required?
Not required for Monitron because it generates ML alarms with its own two-week history.
Monitron provides immediate alarms based on ISO standards for exceeding absolute vibration levels.
Other platforms will typically require a four to eight week sensor history.
6. Key factors to choose assets for sensor mounting.
List and rank assets with a simple scoring system:
Unplanned downtime per year.
Emergency WO rate.
Degradation based failure that can be measured by a vibration and/or temperature sensor.
Above failures are a process of progressive degradation.
Operational profile (must be constant speed +/- 10%)
Asset can be shutdown to install sensor.
Sensor Bluetooth will reach network gateways.
For a DecisionIQ funded Pilot, schedule a pre-visit call for general Q&A and to review asset selection in item 6 above.
7. PdM Training – Hidden secrets and lessons learned from our 24 Pilots:
How to build a PdM program that scales existing technical talent and support budget.
Strategic advantages of a scaled PdM program.
Factors that will build PdM ROI and factors that will reduce it.
How critical assets that never fail will benefit from a PdM program.
Managing Monitron alarms
How Monitron generates alarms.
Managing false alarms (false positives).
How ROI is calculated and documenting wins to support an ROI for the Pilot.
Q&A from attendees.
8. How many sensors for the Pilot?
Sensor count is determined by the historical failure rate. The more failures, the less sensors required to get some wins in the 60-90 term of the Pilot. Overall, a good rule of thumb is 5% of total assets at site for the Pilot. This is feasible for Monitron given its low acquisition and support costs.
Assign sensors to assets. Given that rotating assets require two sensors, our experience shows it best to use a 1.4:1 ratio of sensors to assets.
Order sensors for the Pilot.
9. Install platform (for Monitron users)
Set up AWS account
10. Physical installation of sensors and gateways:
Match sensors to Gateways with hierarchy data.
Verify live connection.
Install sensor Gateways to network.
Install sensors on assets.
11. Checklist for setup completion (DecisionIQ can assist)
Confirm sensor install list is operational.
Confirm personnel are briefed, confirm alarm inspection process in place
Now, Ready for PHASE 2 – Monitoring and Capturing Wins
DecisionIQ on-site meeting agenda will review items 7 thru 10 above:
Strategic objectives: Hear from executive and operations leadership.
PdM training for scalability – Hidden secrets and lessons learned from our 24 Pilots.
Confirm asset selection, assignment of gateways and sensors.
View assets in operation.
Assist with sensor installation (if required).
Monitron Pilot, PHASE 2 – Monitoring and Capturing Wins (60-90 days)
Start 60-90 day monitoring period to demonstrate Monitron performance
These steps are required for monitoring any predictive maintenance system:
1. Alarm monitoring
Continue with standard maintenance practices.
Morning review of each alarm (DecisionIQ can assist).
Use the Monitron dashboard to track and review condition of asset based on its vibration and temperature sensors.
Manage inspections of assets under alarm.
Schedule a periodic inspection of asset to verify its condition using traditional methods, eg, acoustical, infrared, spectral analysis.
There will be a high rate of false alarms in the first month of monitoring as the system learns.
If warranted by inspection, schedule a repair work order during planned downtime.
It is not uncommon for repairs to be scheduled 30 to 90 days after the initial alarm.
2. Documenting results
Capturing wins (successful predictions) and calculating Pilot ROI:
Documenting them for reports to leadership, training other sites.
Calculating downtime, and maintenance savings.
Alarms determined to be false positives and why. Objective is to screen this out. For example:
Asset type.
Sensor installation.
ISO Alarm Class setting (Monitron only)
Operational anomaly, eg, speed change, startup and shut down.
3. Deliverables
Weekly meetings to review alarm status, inspection results.
Short action report to cover meeting.
Monthly reports for project team and leadership.
Include number and type of alarms, WO decisions.
A final, end of Pilot report that compiles monthly reports, plus the following:
Calculate annual ROI for selected assets.
Estimate annual ROI for site-wide deployment across all eligible assets.
Projected budget for deployment.
DecisionIQ can assist with PHASE 2 (above) as follows:
Coaching on weekly calls, interpreting alarms
Assistance with documenting results
Generating deliverables for client
If Monitron is Easy, then Why DecisionIQ?
With 24 Pilots, DecisionIQ will provide complementary experience, skills and hours to maximize results in less time, with greater clarity and a legacy of new skills to scale PdM operations across multiple sites.
Eight examples of how DecisionIQ serves as a catalyst to success in a Pilot:
experienced strategies and plans for achieving high ROI on POCs, Pilots and deployments
Workshops that upgrade skills, achieve positive cultural change, acceptance of AI
Methodologies & tools for managing the false positives inherent with AI-systems
Experience with monitoring data, making decisions to release a work order
Expertise building special tools and supporting platforms
Experience with sensor installation on assets
Experience with building real-world strategic plans for scale
Expertise building reports for training and leadership
Customer review of DecisionIQ Pilot assistance:
Baxter Healthcare worked with DecisionIQ in a highly successful Monitron Pilot that demonstrated an ROI profit of 60x cost. The Pilot was highlighted in a session at the 2022 AWS re:Invent Conference this past November.
"DecisionIQ’s role as a catalyst in our Amazon Monitron Pilot was fundamental to its great success. We achieved an outstanding ROI and the required experience to deploy at scale, which is now underway."
A.K. Karan - Global Senior Director, Digital Transformation at Baxter Healthcare
Baxter International Inc. is an American multinational healthcare company with over $12B in annual revenue and over 40 manufacturing sites globally..
NEXT POST: Asset Selection for Your Monitron Pilot
GetIQ.ai is a blog about building AI solutions for augmenting decision making and empowering people that make them. It's authored by the engineers at DecisionIQ.
ABOUT DECISIONIQ: We are "boots on the ground" factory engineers expert at the adoption of AI and machine learning into operations. As consultants and system integrators we bring our experience in a mix of a well structured programs that enable our clients to produce winning results and maximum ROI. You will learn to use AI to build competitive advantage by increasing plant uptime, quality and yield. With 24 POCs, Pilots and 16 deployments we have achieved for our client’s ROIs as high as 60x and a cumulative $22 million in savings.
AN AWS PARTNER: We provide capability in design of AI solutions for industrial applications. We work with the AWS Industrial IoT services stack, cloud migrations and Amazon Monitron Pilots and deployments. For qualifying customers, many of these capabilities are eligible for AWS subsidized funding.
©Copyright 2023, DecisionIQ, Inc.