Gains:
- Ability to edit blast design parameters (hole diameter, slice, burden, charge) according to the target piece size with AI
- Ability to analyze part size distribution, breakback and vibration/air shock relationship with AI support
- Ability to verify AI recommended blasting design with field measurement, regulatory limit and blasting expert approval
The main way to make rock digpable in open and underground mining is drilling and blasting. A blasting design; It breaks the rock into the desired piece size with parameters such as hole diameter, hole depth, burden (distance between the hole and the free face), slice/spacing, charge (amount of explosive per hole), delay (millisecond ignition timing between holes). Aim; While obtaining the ideal fragmentation for excavation, transportation and crushing, it is to keep back breaking (unwanted very large pieces), dust, air shock, ground vibration and stone blowing within the limits of the legislation. This is a multivariate and straightforward safety-critical optimization problem. AI helps with parameter scenarios, part size analysis and reporting; but the approval of the blasting design belongs to the authorized blasting expert (detonator/responsible) and is verified by field measurement.
Why is part size so important?
Very large chunks (back crushing) means extra crushing, secondary blasting and slow excavation; Very fine particles mean waste of explosives, dust and loss of efficiency in some processes. The ideal size is determined by the downstream crusher's mouth opening and plant capacity. Fragment size distribution is often measured by photo-based image analysis: the heap is photographed, software (increasingly AI-powered) segments the fragments and outputs a size distribution curve (e.g. P80: the size below which 80% of the mass falls). AI can speed up this image analysis; but it is essential to verify scale calibration (there must be a reference scale in the photo) and errors such as missing thin material.
Step by step: AI-powered blast optimization
- Define the goal. Required P80, acceptable breakback rate, regulatory vibration/air shock limit. AI: clarifies the list of goals and constraints.
- Measure the current situation. Photo-based part analysis, vibration measurement, hole deviation. AI: summarizes data.
- Establish a parameter-result relationship. How do particle size and vibration change when burden/spacing/charge changes? AI: relationship analysis and scenario with historical data.
- Suggest and limit scenarios. AI suggests several design variations; each is eliminated according to the regulatory limit.
- Expert approval and trial firing. The selected design is reviewed by the blasting expert and measured with a controlled trial shot.
- Measure, compare, fix. The result is measured, evaluated according to the target, and documented.
Caution: Ground vibration (PPV, mm/s), air shock (dB) and stone flying during blasting are direct risks for nearby structures and people. These limits are set by legislation and any design suggested by AI cannot be implemented without field measurement and authorized expert approval.
Vibration and adjacent structures
Blasting vibration is an effect that decreases with distance but poses a risk to nearby structures. The scaled distance approach is used to estimate the expected PPV by relating distance and charge per hole. AI can explain the logic of this relationship and help build a local prediction curve from your historical measurements; but staying below the limit value requires a model calibrated with site-specific measurement and expert judgment. An overall coefficient given by AI does not reflect the geology of your field.
three mini cases
Case 1 — Reducing backbreaking. In a quarry, 15% of the shots produce large blocks and secondary crushing is required. The team summarizes the parameters and track analysis of the last 30 beats to the AI; They see that large blocks are especially concentrated in pulses in hard vascular zones, reaching 3.2 m above the burden. They try to pull Burden to 2.8 m in these zones; In the test shot, the backbreak decreases to 6%. AI showed the relationship; The expert approved the design, and the measurement confirmed the result.
Case 2 — Vibration limit. There is a village 300 m away from the quarry border. The team feeds past vibration measurements to the AI and has it build a local scale-distance curve; According to this curve, it can be seen that the charge per hole can exceed the limit. The charge per hole is reduced and the delay pattern is changed. The PPV measured in the trial shot remains below the regulatory limit. AI has accelerated calculus; The limit and approval came with expert and measurement.
Case 3 — Scale calibration error (warning). An intern gives the stack photo to the AI and asks for a part size. There was no reference to scale (an object of known size) in the photograph; The AI still "guesses" a P80. The value is half the truth because the scale was unknown. Lesson: scale calibration is essential in photo-based part analysis; A size value without calibration is meaningless and AI can silently generate it.
Copiable prompt templates
BLASTING TARGET AND CONSTRAINT FRAMEWORK "Role: You are assistant drilling and blasting engineer. Help me clarify the target and constraints for a shot: target P80, acceptable refraction, regulatory PPV and air shock limit, adjacent structure distance, deflection tolerance. Make a 'how to measure / who approves' note for each constraint. State that limit values must be verified from the official regulation."
PART SIZE DATA INTERPRETATION "Below are the parameters (burden, spacing, charge) of the final pulses and the measured P80 values. Comment on the observed trend between parameters and part size; mark which parameter seems to be associated with the large part. EXACT design proposal; give hypotheses to be verified by trial shot. Data: [paste]."
VIBRATION SCALE-DISTANCE ANALYSIS "I have historical blast vibration measurements in each row: distance, charge per hole, and measured PPV. Write a code skeleton that fits a local curve using a scaled distance approximation and estimates the expected PPV for a given distance/charge. WARNING: this estimate is site specific and requires regulatory and expert approval. Data: [paste]."
POST-SHOT REPORT DRAFT "Write a post-shot evaluation draft from the following data: design parameters, measured fragment size (P80), backbreak observation, measured PPV/air shock and comparison with regulatory limit, recommendations for improvement. Mark all safety-critical decisions 'subject to authorized blasting expert approval'."
Weak prompt / Strong prompt
WEAK PROMPT: "Give the best blast design: burden, spacing, charge."
STRONG PROMPT: "Role: You are an assistant blasting engineer. Work with my historical blast efficiency and fragment measurements. Analyze the trend between burst/spacing/charge and P80 and backbreak, compare several scenarios to eliminate them according to CONSTRAINTS (legislative PPV, adjacent structure distance). YOU determine the final design; note that each scenario requires trial firing and competent expert approval."
Comparison table: parameter and effect
Parameter
Typical effect when increased
AI role
verification
burden
Bulk/rear break increases
Trend analysis
test shot
charge per hole
Vibration/PPV increases
Scale-distance calculation
Vibration measurement
delay pattern
Fragmentation and vibration vary
Scenario
Measurement + expert
Hole deviation
irregular fragmentation
anomaly sign
Field measurement
specific charge
Fine/powder increases
Comparison
Part analysis
Common mistakes
- Part analysis without calibration. Without a scale reference P80 is meaningless.
- Relying on the overall vibration coefficient. Vibration is site specific; Calibrate with local measurement.
- Implementing the design without trial shots. New parameters are verified by controlled pulse.
- Ignoring the legal limit. PPV, air shock and skidding are human and structure safety.
- Focusing only on part size and forgetting about vibration. Increased charge for good fragmentation could put neighbors at risk.
Tip: Good blasting is always a balancing act: while wanting better fragmentation, don't exceed the limits of vibration, air shock and dust. AI quickly compares scenarios, but it is the competent expert who establishes and confirms the balance.
In summary
Drilling-blasting is the task of keeping vibration, air shock, dust and skidding within the regulatory limits while breaking the rock into the ideal piece size. AI; It is strong in target/constraint framework, parameter-part relationship, scale-distance vibration analysis and shot report drafting. However, part analysis requires scale calibration, vibration estimation requires local measurement; and no design test firing can be implemented without the approval of an authorized blasting expert. Safety limits always take precedence over part size.
Application task
Use the “Part size data interpretation” template with your historical (or sample) shot data to extract which parameter appears to be associated with the bulk. Then set up a local prediction curve from the "Vibration scale-distance analysis" template with your vibration measurements and evaluate whether a distance/charge combination exceeds the limit. Finally, prepare a "Draft post-fire report" for a test shot and mark safety-critical decisions for expert approval.
checklist
- [ ] I verified the scale calibration in the part analysis.
- [ ] I calibrated the vibration estimate with the site-specific measurement.
- [ ] Legislation I have confirmed the PPV/air shock limits from the official source.
- [ ] I validated the new design with a trial shot, I did not apply it directly.
- [ ] I left safety-critical approval to the authorized blasting expert.
- [ ] I evaluated part size and vibration/air shock balance together.