From the lab bench
to the mining bench
ShockAI holds the measured record of each blast: the design, the holes as drilled and loaded, the firing times, VoD, ground vibration and airblast. ShotTrack builds the instruments that record it. We put both to work with universities, research groups, manufacturers and operations studying how blasts behave and how well they can be predicted.
We already work with leading universities, academic groups in machine learning and blast prediction, equipment manufacturers and major operations. Each collaboration is scoped to its question, and to the sites and instruments it needs.
Research is already part of how the platform is built
- Universities and research institutesGroups in mining, geotechnical and civil engineering working on ground vibration, fragmentation and detonation performance, from student projects to funded programmes.
- Machine learning and AI researchersGroups improving blast prediction with learned models: attenuation, frequency content, fragmentation and movement, trained on measured records and checked against the blasts that followed.
- ManufacturersExplosives, initiation-system and equipment makers characterising product performance in the field: measured velocity of detonation, firing-time accuracy and the vibration that results.
- Mining and quarrying operationsSites that host trials on their own benches and take the findings back into blast design and compliance reporting.
Instruments, the record and a platform to run on
Production-grade instruments and data from real blasts are the hard part of blasting research. They are what we can offer.
- Field instrumentationShotTrack velocity-of-detonation recorders and vibration monitors, the same units operations record with, available to research partners for trials and studies.
- The measured blast recordEach blast on the platform carries its design, the holes as drilled and loaded, the firing record, VoD, ground vibration and airblast in one structure. Where an operator agrees, that record is available to a study, de-identified where the work allows.
- A platform to analyse onUpload recordings, work with the analysis tools operations use, and compare predictions with measurements on the same blasts. Write a prediction model to the Model API, the interface our own models run on, and in early access connect it to be held to the same blasts.
- Engineers who fire blastsPeople who design, fire and measure blasts for a living, so a study is scoped around how a site actually operates and a result is judged by whether it would hold on a production bench.
Where we would like to work with you
Six areas where measured field data changes what can be studied. Tell us if yours is not listed.
- Blast vibration predictionAttenuation laws, frequency content and near-field behaviour, and how a site’s own measurements improve its model.
- Learned modelsMachine-learning approaches to prediction and anomaly detection, trained and validated on measured records rather than assumed rock properties.
- Detonation performanceVelocity of detonation in the hole, product and priming behaviour, and what the VoD trace reveals about the charge.
- Initiation timingFiring-time accuracy, its effect on vibration and fragmentation, and what the measured record shows about delay design.
- Fragmentation and movementLinking design and measured detonation to the muckpile: fragmentation, heave and dilution.
- Instrumentation and signal processingSensor placement, sampling and processing for field recordings that have to survive production conditions.
From a conversation to a study
- 01Scope the questionA conversation about what you want to learn, what the instruments and the platform can measure, and what a useful result looks like.
- 02Agree the termsA written agreement covering instruments, data access, confidentiality, publication and intellectual property. Operator data is shared only with that operator’s consent.
- 03Run the studyInstruments in the field, recordings on the platform, and the measured record beside your results as blasts are fired.
- 04Share what was learnedFindings go back to the operations that hosted the work and, as the agreement allows, into publication and into the platform itself.
Start a conversation
A few details about you and what you would like to explore. We reply within two business days.