Countering Technical Failures in Delhi Metro Using Data

Countering technical
failures in Delhi Metro

Using Data Science
MOHIT KUMAR
SR. CRA
19440
9711754512
Problem : Abrupt failure of
equipment
 Delhi Metro has been instrumental in ushering in a new era
in the sphere of mass urban transportation in India.
 But unanticipated technical failures impedes with our
Vision, “Commuting experience in Delhi Metro to be
customer’s delight.”
 Frequent and lasting technical snags not only result in
monetary loss for organization, but also raises reliability and
punctuality concerns.
Solution : Incorporating Predictive
Maintenance along with Preventive
Maintenance
 Predictive maintenance is used for identifying potential failures of assets that
may occur while checking the assets when they are running.
 It uses data analysis tools for controlling the performance and condition of
equipment during normal operation to reduce the probability of failures.
 It is a technique in which the service period of crucial parts is predicted on an
inspection basis so that the parts can be useful within the limit of their service
period.
 Studies have shown that businesses across all sectors of industry can reduce
equipment maintenance costs by 12 to 18% by investing in Predictive
maintenance and tracking failure rates
 Preventive maintenance refers to regular, routine maintenance to
help keep equipment up and running, preventing any unplanned
downtime and expensive costs from unanticipated equipment
failure.
 Preventive maintenance involves a systematic analysis of
equipment where potential malfunctions are detected and
corrected to prevent equipment failure.
 A preventive maintenance schedule may include things such as
cleaning, lubrication, oil changes, adjustments, repairs, inspecting
and replacing parts, and partial or complete overhauls that are
regularly scheduled.
Predictive Maintenance Preventive Maintenance
It is performed to predict failures that might It is performed to prevent assets from
occur unexpected failures.
No downtime of machine is required, i.e. one One needs to increase downtime of the asset
does not need to stop main functions of to carry out maintenance, i.e. one needs to
assets as this maintenance can be performed stop main functions of assets to carry out
while assets are performing their regular maintenance action.
functions.
Maintenance occurs only when potential Maintenance occurs even if potential failures
failures are identified. are not identified.
It is more complex and difficult than It is a less complex process and simple than
preventive maintenance. predictive maintenance.
This maintenance action is less costly than This maintenance action is more costly than
preventive maintenance as one can simply predictive maintenance, as regular
reduce, avoid maintenance that is not maintenance requires more investment.
necessary and thus reduce maintenance
costs.
It is less time-consuming as in this type of It is more time-consuming because in this
maintenance, one needs to perform type of maintenance one needs to perform
inspection and maintenance only when inspection and maintenance regularly.
required.
Why incorporate Predictive
maintenance?
 In any asset-intensive industry, reactive maintenance can interfere with
operational efficiency and profit margins.
 Incorporating Predictive maintenance along with Preventative
maintenance can yield significant financial and organizational benefits.
 Studies have shown that businesses across all sectors of industry can
reduce equipment maintenance costs by 12 to 18% by investing in
Predictive maintenance and tracking failure rates.
How to implement?
• The process starts with identifying the problems to solve. The
problem should be broad and broken down into specifics. Identify Problems
• Once the problems are identified, the data needs to be

collected for analysis. Identify data
• If the data will not be available, in such case, infrastructure to required
collect data needs to be built.
• While building the infrastructure and processes, attempt can be Build infrastructure
made to enhance the utility of such infrastructure/processes. to collect data
This can be done by assessing other possible utilities of the

data collected. Collect , Clean
• If the marginal cost of adding more data to solve a significant and visualize data
problem is low, it can be pursued.

• Once the data starts getting collected, it needs to be cleaned Build and test
and visualized. models
• The dashboards and visualizations are to be followed by the

creation of predictive model(s), which are to be tested, Review and
reviewed and deployed. deploy

Failure data collection
There are 5 key metrics to failure data collection.

 Mean Time To Repair (MTTR)
 Mean Time Between Failure (MTBF)
 Mean Time to Failure (MTTF)
 Planned maintenance percentage (PPC)
 Planned maintenance compliance (PMC)
Mean Time To Repair (MTTR)
 MTTR , Mean Time To Repair is a measure of the average time that elapses
between the discovery of equipment breakdown and the equipment's return to
operational status.
 This includes the time it takes to troubleshoot the issue, the time spent in actual
repairs, and the time spent in testing before the machine or piece of equipment
is utilized again.
 It can be an invaluable aid in the decision making process regarding the repair vs
replace question, maintaining inventory levels of replacement parts and tools and
staffing levels.
Mean Time Between Failure (MTBF)
 MTBF Mean Time Between Failure gives an indication not only of

the availability for each piece of operational equipment, but also the level
of reliability.
 MTBF measures the average elapsed time between equipment breakdowns.
 We can calculate MTBF by taking the total operational time of the machine in
question and dividing that figure by the number of failures within that time
frame.
 While MTBF is not directly related to planned maintenance, it is extremely
helpful for determining the optimal frequency for certain tasks, such as
equipment inspections, lubrication, etc.
 MTBF works with repairable equipment.
Mean Time to Failure (MTTF)
 MTTF is a measure of the expected length of time a particular machine or piece

of equipment will stay operational.
 MTTF must be calculated as an average of multiple non-reparable units.
 Effective maintenance and using durable, high quality components in pieces of
production equipment can boost MTTF.
 It can give a good idea of how long components will last, which in turn can help
them to plan for downtime, keep a sufficient number of replacement parts on
hand, and analyze the pros and cons of sourcing the components in question
from a different supplier.
 MTTF focuses on non-reparable items.
Planned maintenance percentage (PPC)
 Planned maintenance percentage gives a proportion of time spent on planned

maintenance compared with total maintenance time.
 If PPC numbers are sub-optimal, however, a detailed analysis of which systems
are suffering the majority of unexpected failures can help managers prioritize
areas of need, and implement strategies to mitigate the effect of unplanned
downtime.
 While PPC may be a broad metric, it can give a solid indication of how effective
current preventative maintenance program actually is.
Planned maintenance compliance (PMC)
 Planned maintenance compliance is a measure of how well maintenance

department is performing their required tasks.
 It can give a percentage of completed work orders compared to total work orders
requisitioned.
 PMC is helpful in identifying maintenance staff's performance, and can also help
you to determine if staffing levels should be adjusted, or additional training
should be provided in certain areas.
Conclusion
 Since Preventive maintenance is not always needed, it could be

unnecessary time spent on a job.
 Our organization already has infrastructure in place for data gathering,
therefore, no new infrastructure is needed.
 With training, existing staff can be used to clean and visualize data at hand.
 Inventory can be managed more efficiently by predictive analysis of
components which are likely to fail.
 Preventive and predictive maintenance needs to be carried out as both
techniques are the two sides of the same coin but different altogether.
Thank you

Countering Technical Failures in Delhi Metro Using Data

Uploaded by

Copyright:

Available Formats

Countering Technical Failures in Delhi Metro Using Data

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Countering Technical Failures in Delhi Metro Using Data

Uploaded by

Copyright:

Available Formats

Countering technical

failures in Delhi Metro

• Once the problems are identified, the data needs to be

This can be done by assessing other possible utilities of the

problem is low, it can be pursued.

• The dashboards and visualizations are to be followed by the

reviewed and deployed. deploy

There are 5 key metrics to failure data collection.

 MTBF Mean Time Between Failure gives an indication not only of

 MTTF is a measure of the expected length of time a particular machine or piece

 Planned maintenance percentage gives a proportion of time spent on planned

 Planned maintenance compliance is a measure of how well maintenance

 Since Preventive maintenance is not always needed, it could be

You might also like