Zero downtime - Is it really possible?

Zero downtime - Is it really possible?

19 July 2016

Comments:

0
 July 19, 2016
 0
Category: Big Data

Businesses are – in general – meant to maximize revenue and profit.  To do that, they need to have a fully operational capacity. Operational downtimes have negative effects and are bad for business.

Do we know really the cost of downtime?

If you are an airplane carrier you need your planes to be up in the air. If you are a green energy company, you need your windmills turning. If you manufacture cars, the machines and robots in your factory have to be always operational. Any downtime will have tremendous costs.

Two type of costs will impact the business in case of downtime. Tangible costs and intangible costs. Tangible costs are the ones you can quantify easily: loss of  inventory, wages, lost capacity, lost production, the cost of operations etc.. Intangible costs are the ones not directly quantifiable: impact on reputation, employee morale, stress, loss of innovation momentum etc…

A report from Dun & Bradstreet stipulates that 59% of Fortune 500 companies are experiencing a minimum of 1.6 hours of downtime every week. For a company with 10,000 employees, this means a yearly cost of $46 million only in wages.

So what are today our options to avoid downtime?

The report from Dun & Bradstreet indicates that the three industries that are the most impacted by downtimes are the Energy, Telecom and Manufacturing industries.

Trying to combat downtime is not new. CMMS software are often used to track and monitor the machine downtimes.Many methods already exist and have been used for many years:

1- Pre-scheduled regular maintenance

It is like your car really. Your car manufacturer will tell you to do a maintenance every 10,000 miles and would void the warranty if you don’t follow the recommandation. In the same way, regular maintenance operations are done today in all industries. These are often scheduled by the manufacturer or regulated by law. For example in the aviation industry, these checks and maintenance are mandatory. The scheduling of the maintenance can also be linked to the number of work orders and can be tracked by a CMMS.

2- Logging and keeping track

Sounds obvious but logging and keeping track of the downtimes will help understand and combat them. You can’t solve a problem if you don’t understand.

3- Sensor tracked simple diagnostics

Some sensors can track the vibrations and the stress put on the machines. By analyzing the information collected, some abnormal behaviour can be identified and maintenance operations scheduled.

Nevertheless, all these solutions won’t put an end to panic fixes. Failures and outages will arise if we can’t totally understand everything linked to the in-service equipment.

 

And what about tomorrow, how can technology help us combat downtime?

 

Well, there are new solutions that can prevent downtimes by predicting them before they arise. These solutions take advantage of new technical improvements, uses Big Data and Machine Learning. That’s what we call Predictive Maintenance.

 

The main idea behind predictive maintenance is to predict when to do a maintenance operation based on the condition of the machine. More than analysing a few available information, we can concentrate on gathering all machine information, study the insignificant variations of data and analyse the imperceptible corrélations between them.

 

The reason is that the reality on the ground is often complex. To solve a problem before the failure happens, we need to combine all available information which is in general low signal information and alone bring only little value. By combining them we can we create a gold mine that can help predict outages and downtime.

 

Leave a Reply

Your email address will not be published. Required fields are marked *