Real-Time OEE Analytics for Drilling Rigs

By David Cook on October 6, 2026

real-time-oee-analytics-drilling-rigs

A rig can be on location, crewed and turning for 30 days and still give away most of a week. Some of that is downtime everyone sees. More of it hides in routine work — connections, trips, slow footage — done a little slower than the rig's own best. The morning report shows the first kind and misses the second, so the same hours are lost again on the next well. iFactory's OEE Analytics Suite reads the rig's sensor data in real time and turns it into one score with three reasons behind it, by rig, pad, crew and shift. To see it on your own rig data, book a rig review.

Upstream Drilling · OEE Analytics Suite

Real-Time OEE for Drilling Rigs: See Where Every Rig Hour Goes

One live score for availability, performance and quality — compared across pads, crews and shifts, so you can cut NPT and lift metres per day.

  • Built from signals the rig already records
  • Downtime and slow routine work shown apart
  • Fair benchmarks, not a blame chart
Rig 14 · last 30 daysillustrative
76.4%OEE
Availability90.0%
Performance87.5%
Quality97.0%
Three good-looking numbers multiply into one that is not: 151 of 640 planned hours went somewhere other than making hole.
The page in 20 seconds
  • OEE turns a rig's month into one number and three reasons.
  • The biggest loss is often not downtime. It is routine work done slower than your own best.
  • Fair benchmarks across pads, crews and shifts show where to look first.
598rigs running in the US on 2 October 2026, by the Baker Hughes count
6.6–13.2 minspread in average drilling connection time between operators in a 170-well Bakken study
15%faster well delivery after one operator in that study tracked connection and tripping KPIs
16% → 9%NPT cut by one offshore operator in about a year, as reported by McKinsey

Rig OEE in Plain Terms: One Score, Three Questions

OEE asks three things of every rig hour: was the rig working, was it working at pace, and did the work stick?

The idea comes from manufacturing. It fits drilling because a rig costs the same each day whether or not it makes hole. Multiply the three answers and you get one number that crews, superintendents and managers all read the same way. Our support team can help fit the definitions to your operation.

It matters more each year. The US Lower 48 rig count fell from 750 in December 2022 to 517 in October 2025 while oil and gas output reached record highs, according to the EIA. Fewer rigs are doing more work, so every rig hour is watched more closely.

A

Availability

Was the rig working?

Sum: hours worked ÷ hours planned.

What eats it: equipment repairs, waiting on weather or services, stuck pipe, lost circulation.

P

Performance

Was it working at pace?

Sum: benchmark hours for the work done ÷ hours it took.

What eats it: slow connections, slow trips, footage below the pace of the offset well.

Q

Quality

Did the work stick?

Sum: metres right first time ÷ metres drilled.

What eats it: reaming tight hole, re-drilling, sidetracks, a casing run that needs a second attempt.

Worth knowing

Drilling has no single standard for OEE. Some teams leave quality out. Others count rig moves as lost time. Neither is wrong — but pick one definition and keep it, or your rigs cannot be compared.

Where 640 Planned Hours Went

One rig, 30 days, two wells on a pad. Here is the whole month in a single bar.

The rig had 720 calendar hours. Take away 80 planned hours for the rig move and scheduled maintenance, and 640 were there to be used. The figures are illustrative. To see the same bar built from your own data, book a working session.

Productive — making hole at pace, right first time489 h
Rework — reaming and re-drilling15 h
Invisible lost time — working, but slower than benchmark72 h
NPT — stopped for something unplanned64 h
Factor
In plain words
The sum
Result
Availability
Hours worked, out of hours planned
(640 − 64) ÷ 640
90.0%
Performance
Benchmark hours, out of hours worked
504 ÷ 576
87.5%
Quality
Metres right first time, out of metres drilled
11,640 ÷ 12,000
97.0%
OEE
All three multiplied
0.900 × 0.875 × 0.970
76.4%
What about metres per day?

Metres per day is the result; OEE is the explanation. This rig averaged 400 metres a day. On the same wells at 100% OEE it would have made about 524. The score tells you which of the three losses holds the difference.

Read it this way: the rig was fully productive for 489 of 640 hours. The other 151 hours — more than six days — went to three different kinds of loss. Each one needs a different fix, which is why a single downtime figure is not enough.

The Loss You Report and the Loss You Don't

Downtime gets a line in the morning report. Slow routine work does not — and on this rig it is the bigger number.

NPT is when the operation stops for something unplanned. Invisible lost time is different: the rig is working, just slower than a pace it has already shown it can hit. In a published study of 170 Bakken wells, average drilling connection time ranged from 6.6 to 13.2 minutes between operators. Ask our drilling specialists how a benchmark is set.

NPT

64 h

Seen, coded and reported.

Rig equipment repair22 h
Downhole tool failure16 h
Waiting on weather, services, orders14 h
Hole problems: stuck pipe, losses12 h

Invisible lost time

72 h

Never stopped, so never reported.

On-bottom pace below the offset well24 h
Tripping in and out22 h
Drilling connections18 h
Flat time: casing runs, pressure tests8 h
420drilling connections in the month
× 2.6 minaverage time over benchmark
= 18 hgone, a few minutes at a time

Small gaps add up because they repeat. In the Bakken work, one operator brought connection time from 9.9 to 6.7 minutes and well delivery from 15.4 days to about 13 — by measuring routine steps crew by crew and standardising the best practice.

Where the hours usually come back first

1

Connections

Agree who does what before the stand is down. Time each step of the connection; one of them is usually far longer than the rest.

2

Trips

Track speed per stand in cased hole, where going faster carries little risk. Open hole keeps its own, slower limit.

3

Repeat faults

A top drive or mud pump fault that returns every few weeks is a maintenance job, not bad luck. Count repeats, not only hours.

4

Waiting

Hours spent waiting on tools, trucks or decisions are planning problems. They stand out once every wait carries a reason code.

What 52 Recovered Hours Are Worth

Nobody recovers every lost hour. Take half of the invisible lost time and a quarter of the NPT on this one rig, and the month looks like this. Swap in your own spread rate.

One rig, one monthillustrative
Half the invisible lost time36 h
A quarter of the NPT16 h
Hours back52 h
Value at $3,000 an hour$156,000
Metres per day400 → 432
Assumes a $72,000-a-day spread rate for rig and services, and that recovered hours drill at the month's average pace.

Benchmark Pads, Crews and Shifts — Fairly

The same rig can look like two different rigs between day and night. A fair comparison shows which gaps are real.

iFactory lines up pads, crews and shifts on the same measures and the same kind of work. The aim is not a league of winners and losers. It is to find the best practice already on your location and spread it. To see your own crews side by side, book a benchmark review.

Crew
Shift
Connections
Average time
Over benchmark
Hours lost
Crew A
Day
105
7.9 min
+1.2 min
2.1 h
Crew B
Night
105
10.1 min
+3.4 min
6.0 h
Crew C
Day
105
8.6 min
+1.9 min
3.3 h
Crew D
Night
105
10.6 min
+3.9 min
6.8 h
All crews
—
420
9.3 min
+2.6 min
18.2 h

Illustrative. Drilling connections, weight-to-weight. The 6.7-minute benchmark is the figure reached in the published Bakken study; yours should come from your own best wells.

Look at the pattern, not the ranking. Both night crews are about two minutes slower per connection than both day crews. That points at the night, not at either crew — and closing that one gap would return about seven hours a month.

OEE by pad

illustrative

Same rig design, same month.

North pad81%
Central pad76%
South pad71%

Ten points between pads is a question, not a verdict: geology, well design and logistics all differ.

Four rules for a fair comparison

  • Same hole section. Surface, intermediate and lateral work are scored apart.
  • Same kind of rig. Pipe-handling equipment changes what is possible.
  • Enough samples. Judge a crew on dozens of connections, not one bad stand.
  • Ask why before who. A night gap often traces to lighting, staffing or the section being drilled.

How the Score Stays Live

No stopwatch and no extra reporting. The score is built from signals the rig already records every second.

The rig's drilling recorder logs hookload, block height, rotary speed, torque, pump pressure and depth. iFactory reads that stream, works out what the rig is doing and times it. Published work from a major service company reports about 15% potential time saving on connections once they are measured automatically. Our integration team can confirm which signals your rigs expose.

1

Read

The sensor stream comes in from the rig, with the daily drilling report beside it.

2

Recognise

Each second is tagged: drilling, connection, tripping, circulating, reaming or waiting.

3

Time

Every activity is timed against the benchmark for that hole section. Stops are matched to NPT codes.

4

Score

OEE refreshes by rig, pad, crew and shift, with the top losses listed in hours.

Who sees what

  • Driller. A connection timer against benchmark at the end of each stand.
  • Superintendent. Yesterday's score and the three biggest losses, every morning.
  • Drilling manager. Rigs and pads ranked, with the trend over the last wells.
  • Everyone. The same definitions, so the morning call argues about fixes, not figures.
Example exchange · illustrative
SuperintendentWhy is Rig 14 behind plan this week?
iFactory AIRig 14 is at 71% OEE this week against 76% for the month. Availability cost most: 9 hours of NPT on the top drive on Tuesday night. Night-crew connections averaged 10.4 minutes against a 6.7 benchmark — about 3 hours. Quality was clean, with no reaming.
SuperintendentWhich fix is worth more?
iFactory AIThe top drive fault has now occurred twice in 30 days; a work order is drafted for your approval. Bringing night connections to the day crew's 7.9 minutes would return about 2 hours a week.

Turnkey AI: Delivered, Connected and Live in 6–12 Weeks

You do not build this. It arrives ready.

iFactory ships as a pre-configured NVIDIA AI server, racked and ready, with the software pre-loaded. Rack it, plug in power and Ethernet, and the AI is live on your network.

Our team handles cabling, network setup, PLC and SCADA integration, operator training and 24×7 remote monitoring. The server sits inside your own network, and your drilling data stays on it. For a scope matched to your rig fleet, request a turnkey quote.

Weeks 1–4

Ship, network and data

Server installed at the rig or in your operations centre. Sensor stream and daily reports connected. OEE definitions and NPT codes agreed.

Weeks 5–8

Model training and pilot

Rig-state detection tuned on your own wells. Benchmarks set by hole section. One rig scored in shadow and checked against the reports.

Weeks 9–12

Go-live and training

Live dashboards for drillers, superintendents and managers. Training by role. 24×7 remote monitoring begins.

Live in 6–12 weeksfrom delivery to crews using it
1000+ clientsacross industrial operations
99.9% uptimewith 24×7 remote monitoring

Frequently Asked Questions

What is OEE for a drilling rig?

It is one score made of three parts: availability (was the rig working), performance (was it working at benchmark pace) and quality (was the hole right first time). Multiply them together. The result shows how much of the planned rig time was fully productive.

What is the difference between NPT and invisible lost time?

NPT is a stop: a repair, a wait, a stuck pipe. It is coded in the daily report. Invisible lost time is routine work done slower than a known benchmark, such as a slow connection. Nothing stops, so nothing is reported — unless every activity is timed.

What is a good OEE for a rig?

There is no industry figure worth quoting, because companies define the parts differently. The useful comparison is against your own best rigs and wells, on the same definitions. Watch the trend and the size of each loss, not the headline number alone. A rig that moves from 72% to 78% on steady definitions has gained real days.

What data do you need from the rig?

The sensor stream the rig already records — hookload, block height, rotary speed, torque, pump pressure and depth — plus the daily drilling reports. In most cases no new sensors are needed. Well plans and offset well data help set fair benchmarks.

How do you compare crews without it turning into blame?

By comparing like with like: the same hole section, the same rig type and enough connections to be fair. The gap is shown in hours and by step, so the talk is about method. Many gaps turn out to be conditions, not people.

Does it work on land rigs and offshore rigs?

Yes. The three questions are the same. What changes is the planned-time rule, the benchmarks and the value of an hour, which is far higher offshore. The example on this page is a land rig drilling pad wells.

How long does it take to go live?

Six to twelve weeks from delivery. We need a place for the server with power and Ethernet, access to the rig's sensor stream and your daily reports. A pilot normally covers one rig. To check your set-up first, contact our team.

Bring One Rig and One Month of Data

In thirty minutes we show how the month would score, where the hours went and which loss is worth chasing first. You keep the picture whether or not you go further with iFactory.

Five things worth bringingif you have them
  • 1A month of sensor data for one rig
  • 2The daily drilling reports for the same month
  • 3Your NPT code list
  • 4Well plans or offset wells to benchmark against
  • 5The crew and shift pattern

Share This Story, Choose Your Platform!