Analyze machine histories, compare production cycles and prepare the data behind quality and maintenance models. l brings qSQL and numerical array operations to recorded sensor measurements, batch results and operating conditions.
01 / The workload
Cycle data, sensor channels, batch outcomes and maintenance records describe related parts of a process. Engineers need to compare conditions and inspect the interval around a change. Consistent calculations and clear identifiers make those comparisons useful across repeated investigations.
Filter recorded measurements by asset, channel and interval. Calculate ranges and summary statistics, then compare the cycle or event window with a reference population.
Group batch and test outcomes by recipe, production line and equipment state. Compare distributions and retain the records needed to investigate a change in quality.
Transform recorded condition measurements into numerical features for your existing maintenance models. Compare feature behavior against the asset events and labels you supply.
Analyze recorded state changes, event durations and production counters. Break down an operational summary to understand which intervals and assets contribute to it.
02 / From data to analysis
A cycle-level analytical dataset gives quality and reliability work a consistent starting point while retaining the underlying samples.
Bring recorded measurements with asset identifiers, cycle boundaries and the operating context required for the comparison.
Select the intervals, derive numerical columns and group the measures by asset, cycle or production condition.
Compare populations and send the relevant features or sample extracts to your existing engineering and modeling tools.
/ Compare recorded cycles l>t:([]cycle:1 1 2 2;temp:70 74 68 80) l>select avg temp by cycle from t cycle| temp -----|----- 1 | 72 2 | 74 l>select max temp by cycle from t cycle| temp -----|----- 1 | 74 2 | 80 l>exec cycle from t where temp>75 ,2 l>
03 / Why l
You write q/k. The runtime considers the physical representation of the data, choosing encoding-aware operations where supported and materializing values when needed.
04 / Your first evaluation
Choose a completed production run with cycle boundaries, related sensor channels and a known change in process behavior.
Available as a research preview for local evaluation.
Why moving fewer bytes matters for analytical workloads.
Read the paper Integration guideCall l from Python, pass typed values and bring results back to your application.
Read the guide Integration guideBring flat columnar datasets into l through the native Parquet reader and writer.
Read the guideYour data. Your next question.