Consumption outside operating hours
A load remains high overnight, at weekends or during a shutdown.
Data investigation for industrial sites
We find the unexplained patterns in your data that deserve attention.
AGAINWARD investigates your existing energy, production and operational data to identify unusual consumption, gradual drifts and unexplained patterns. Where the evidence allows, we quantify the associated energy and cost.
You see which deviations may be affecting your costs, what the data actually proves and where to investigate first — without deploying another platform.
aurele.againward@gmail.com01Uses data you already have
02Quantifies energy and cost when the data allows
03Clear reporting with explicit limitations
A concrete example · synthetic data
While idle, the site is using almost twice as much power as usual.
AGAINWARD identifies the deviation across 15 nights, measures the additional energy and estimates its cost. The question for the site team is whether this consumption is operationally necessary.
No production. Sustained consumption.
H01 validation case · no client results
01 / What the data shows
The same idle conditions.
A different level of consumption.
Usual overnight level22 kW
Observed overnight level40 kW
+18 kW
The highlighted section is the deviation to investigate.
15 out of 15 nights
The pattern is consistent. Production remained at zero throughout the 120 overnight hours analysed.
02 / What this deviation represents
Energy associated with the observed deviation2,152.8 kWh
Total across the 15 nights investigated.
Estimated associated cost€376.74
Using the demonstration tariff of €0.175/kWh.
The avoidable share still needs to be established.
This is the cost associated with the observed deviation, not a guaranteed saving. The data does not identify the cause or the equipment responsible.
03 / The next check
Identify which equipment must remain active overnight, then establish whether this load serves an operational need or reflects a setting that should be changed.
How it works
Three steps, with no new tool to deploy.
The initial discussion establishes whether your question and available data are suitable for a pilot investigation. No real operational data is requested before an agreement is in place.
- 01
You describe the case
The relevant scope, the issue you have observed and the type of export available. No real files are needed for the initial discussion.
- 02
AGAINWARD investigates your exports
Once agreed, we validate the data and investigate anomalies, drifts, recurring patterns and competing explanations.
- 03
You receive the findings
A prioritised report sets out the evidence, hypotheses, limitations and the checks to carry out next.
What we investigate
Concrete issues in their operational context.
A signal is only retained after comparison with actual activity, operating hours and the other operational context available.
Gradual drift
Consumption rises over time without a corresponding change in activity.
Declining performance
Comparable production requires more energy than it did previously.
Recurring cycle or peak
An unusual pattern returns at specific times or under specific conditions.
A false lead to rule out
Weather, a production change or a particular schedule explains the observed deviation.
Possible data sources
EnergyTimestamped power, energy or meter readings
ActivityProduction, cycles, batches or running states
ContextSchedules, shutdowns, maintenance, temperature or tariff
We work from files you already have, usually CSV or Excel exports. An investigation does not require every source listed here.
AGAINWARDInvestigation report
What the supplied
data shows
data shows
- 01Scope and data quality
- 02Main observed evidence
- 03Hypotheses and competing explanations
- 04Limits of the findings
- 05Recommended checks
Written for management, production, energy and maintenance teams.
What you receive
A report you can use.
No additional interface.
The deliverable brings together the evidence that matters when deciding where to look and which check to run.
- The main signals, supported by the necessary charts.
- Their scale, frequency and context.
- A clear distinction between evidence and hypotheses.
- Data limitations and explanations that were ruled out.
- The next checks, ranked by usefulness.
A concise PDF report, usually 3 to 8 pages depending on the case.
Method and limitations
We do not attribute a cause that the data cannot prove.
A pattern can be measured even when the responsible equipment cannot be identified. In that situation, the report says so and recommends the minimum check needed to investigate further.
1Observe
→
Measure the deviation precisely.
2Compare
→
Check activity, schedules and context.
3Test
→
Investigate competing explanations.
4Conclude
Stop at the level the evidence supports.
What the investigation can establish
- A measured change and when it occurred.
- Whether a pattern is recurring.
- An energy deviation and its associated cost when the tariff is known.
- Which hypotheses are consistent with or contradicted by the evidence.
What it does not invent
- A fault or physical mechanism that was not observed.
- A responsible asset without independent evidence.
- A recoverable or guaranteed saving.
- Precision that the data quality cannot support.
Direct contact
Do you have an export and a specific concern?
Simply describe the relevant scope, what you have observed and the type of operational or energy data available. Do not attach any real data to your first message.
aurele.againward@gmail.comYou will receive an email response on whether the case is suitable before any file is transferred.