After the crowd crush incidents that have made international headlines over the past decade, crowd management technology has received enormous investment. AI-powered systems that monitor crowd density in real time, predict dangerous buildups before they happen, and provide actionable guidance to event managers are now commercially available from multiple vendors.

So why aren’t more Australian music festivals using them?

I’ve spent the past few weeks talking to festival organisers, crowd management specialists, and technology vendors about the gap between what’s technically possible and what’s actually being deployed at Australian events. The answers are more nuanced than “festivals are cheap” or “organisers don’t care about safety.”

What AI Crowd Management Tools Do

The technology has evolved substantially from simple headcount systems. Current AI crowd management platforms combine several capabilities.

Density Monitoring

Camera-based systems using computer vision can estimate crowd density across an event site in real time. The systems don’t identify individuals — they measure the number of people per square metre across different zones, producing density heatmaps that update continuously.

The critical threshold that crowd safety experts monitor is around 4-5 people per square metre. At this density, individuals lose the ability to move independently and crowd dynamics become dangerous. The AI systems can detect when density in any zone approaches this threshold and alert event managers before a dangerous situation develops.

Crowd Connected and Riskpal are among the vendors offering these capabilities specifically for the live events industry.

Flow Analysis

Beyond static density measurement, AI systems can analyse the direction and speed of crowd movement. This matters because crowd safety incidents typically involve not just high density but also conflicting flow patterns — large numbers of people trying to move in different directions in the same space.

Flow analysis can identify bottlenecks where conflicting flows create pressure points, detect sudden changes in movement patterns that might indicate a panic response, and predict how crowds will move based on historical patterns and current conditions.

Predictive Modelling

The most advanced systems combine real-time monitoring with predictive models that forecast how crowd conditions will evolve over the next 15-60 minutes. These predictions account for scheduled events (a headliner starting on the main stage will draw crowd movement toward that stage), weather conditions, and observed patterns from earlier in the event.

This predictive capability gives event managers time to respond proactively rather than reactively. Opening additional gates, redirecting wayfinding, or temporarily holding entry to a high-density area are all more effective when done before a problem develops rather than after.

Why Adoption Is Slow

Cost

This is the most frequently cited barrier, and it’s real. A comprehensive AI crowd management system for a major festival — including camera infrastructure, edge computing hardware, software licensing, and specialist operators — can cost $150,000-400,000 per event.

For a major festival with revenue of $20-50 million, this is a manageable expense. For mid-sized festivals operating on tight margins, it’s a significant cost that competes with other safety investments like medical services, security staff, and infrastructure.

The counter-argument from technology vendors is that the cost of a crowd safety incident — in human terms, legal liability, insurance costs, and reputational damage — far exceeds the cost of prevention technology. This is true, but festival organisers working with finite budgets still have to make difficult allocation decisions.

Infrastructure Requirements

Most Australian festival sites are temporary installations in parks, rural properties, or multi-use venues that don’t have permanent camera or network infrastructure. Installing the camera and computing infrastructure needed for AI crowd monitoring requires significant setup time and technical capability.

Festivals in permanent venues or established sites have an advantage here because infrastructure can be installed once and reused each year. Pop-up festivals in new locations face much higher infrastructure costs.

Skills and Integration

AI crowd management tools produce data and alerts, but someone needs to interpret that information and make operational decisions. Festival crowd management is traditionally a role filled by experienced event managers who rely on visual observation, radio communication, and instinct developed over years of experience.

Integrating AI-generated data into existing crowd management workflows requires training, practice, and cultural change. Some experienced crowd managers are sceptical of technology that they feel duplicates capabilities they already have. Others are enthusiastic but need time to learn how to incorporate AI data alongside their existing methods.

Where It’s Working in Australia

A few Australian events have adopted AI crowd management tools and are demonstrating what’s possible.

Major stadium events in Sydney and Melbourne have been early adopters, partly because the permanent venue infrastructure makes installation easier and partly because the venue operators have the budgets and institutional capability to support the technology.

Live Performance Australia, the peak body for the live entertainment industry, has been advocating for broader technology adoption in crowd management and has facilitated several pilot programs at major events.

One festival organiser I spoke with piloted a camera-based density monitoring system at their 2025 event and described it as “like getting glasses for the first time.” They could see crowd dynamics across the entire site simultaneously rather than relying on reports from scattered ground staff. They identified two density buildups during the headliner set that their ground staff hadn’t reported, and were able to open overflow areas before the situation became dangerous.

The pilot cost them $95,000 including equipment hire, installation, software, and specialist operators. They’ve committed to using the system again this year and are exploring whether they can share infrastructure costs with other festivals using the same site.

The Data Privacy Question

Camera-based crowd monitoring raises legitimate privacy concerns. Even systems designed to count people rather than identify them are collecting video of festival attendees, and the distinction between counting and identification is a matter of software configuration rather than hardware capability.

Event organisers deploying these systems need to address privacy through clear notice to attendees, data retention policies that specify when footage is deleted, technical measures that prevent re-identification of individuals from density data, and compliance with Australian privacy law.

The privacy challenge is manageable but requires deliberate attention. Organisers who treat it as an afterthought risk both regulatory issues and public backlash.

What Needs to Happen

The technology is ready. The barriers are practical — cost, infrastructure, skills — rather than technical. Overcoming them requires a few things to happen.

Shared infrastructure models, where multiple festivals using the same site share the cost of permanent installation, can reduce per-event costs significantly. Companies like Team400 are working with event organisers to build shared analytics platforms that reduce the per-event technology cost.

Industry training programs that teach event managers how to work with AI crowd data alongside their existing skills would accelerate adoption. This isn’t about replacing experienced crowd managers — it’s about giving them better tools.

Regulatory expectations may ultimately drive adoption. As AI crowd management technology becomes more established and more affordable, it’s likely that licensing authorities and insurers will begin expecting its use at large events. This has already happened in some overseas jurisdictions.

Crowd safety is not optional, and the technology to improve it is available now. The Australian live events industry has a strong safety record, but it can be better. AI crowd management tools are a practical, proven way to make events safer for everyone.