Synchronizing Multiple Devices in an Installation
Contemporary installations often behave like distributed instruments: a wall display, sensor, visitor’s phone and lighting controller may all respond to the same gesture. Their agreement can look effortless, yet each device runs its own operating system, timer and network stack. The difficult work is making separate systems share a dependable sense of time, state and sequence.
Reliability becomes especially visible when installation teams move between prototypes and public venues. A working signal in a studio may encounter congestion, delayed packets or restricted background activity in an exhibition. Good device synchronization therefore combines accurate clocks with graceful recovery, giving every component a clear view of what has happened and what should happen next.
Clocks Drift Before Networks Fail
Clock drift begins whenever two devices measure elapsed time independently. Their internal oscillators diverge gradually, while network latency and jitter introduce more variation. Even a small mismatch can cause a light sequence to lag behind sound, or a visitor’s selection to appear after the screen has moved on. Using a monotonic clock for local intervals reduces some errors, but it does not solve coordination across separate machines.
A dependable architecture needs one authoritative time source and a clear state model. Precise Time Protocol can support clock alignment on suitable wired networks, while NTP or application-level timestamps provide more practical options for many installations. Sequence numbers, heartbeat messages and acknowledgement rules add the context needed to distinguish a delayed event from one that never arrived.
This experimentation fits the creative scope of the About ITP 30 Show page, where artists explore how technology changes participation. Synchronisation is rarely the visible subject of an installation, yet unreliable timing can interrupt the relationship between artwork and audience.
Designing for Australian Network Conditions
Public venues in Sydney, Melbourne, Brisbane and other dense cities rarely offer a predictable wireless environment. Apartment blocks, concrete walls, busy exhibition floors and large audiences can crowd the 2.4 GHz band. Using less congested Wi-Fi channels, prioritised access points and wired Ethernet for stationary equipment can reduce contention, but visitors’ phones may still connect through a congested channel.
Captive portals and variable mobile coverage create another layer of difficulty. A person moving through an underground station on the Sydney Metro or Melbourne tram may briefly lose connectivity, while a gallery’s Wi-Fi may depend on a separate internet service. The experience should continue locally when the cloud becomes unavailable, then reconcile its state when connectivity returns instead of discarding the visitor’s interaction.
Time-zone and daylight-saving changes also deserve explicit testing. Sydney, Melbourne, Adelaide and Hobart shift their clocks during daylight-saving periods, while Perth and Brisbane do not. Store event timestamps in UTC, record the intended local presentation time separately and test transitions near the affected dates so scheduled content does not jump or arrive an hour late.
Coordination Needs Observable State
Event messages should describe intentions and state changes rather than rely on every device taking action at precisely the same instant. A lighting command can carry a target time, sequence number and event identifier. Receivers can then decide whether to act immediately, wait briefly or ignore an event that has already been handled. This approach creates tolerance for variable latency without hiding genuine errors.
Browser restrictions, screen locks and mobile operating systems can interrupt delivery. A mixed Australian audience using iOS and Android devices may encounter different background limits, notification behaviour and power-saving settings. WebSockets, MQTT, OSC or vendor APIs can all work, but each requires health checks, reconnection logic and a defined fallback. A local control service often provides a safer core than a cloud-dependent coordinator.
The design sprint account behind a popular exhibit offers a useful parallel: a rapid prototype can reveal where complexity belongs. A central dashboard, synthetic event generator or simple debug view lets technical collaborators test timing, delayed devices and duplicate messages before visual polish obscures the cause.
Rehearse the Failure States
A successful rehearsal covers more than the intended visitor journey. Teams should inject packet loss, reordering, clock skew, intermittent Wi-Fi, browser suspension and sudden power loss. Logs should include device identifiers, message timestamps, applied state and connection changes. These records allow an operator to determine whether a fault came from the artwork, network infrastructure or a particular device.
Testing should also include realistic public behaviour. Visitors may unlock phones to photograph the installation, move between rooms with one hand occupied or enter a code on a screen before the artwork begins. Australian Privacy Principles under the Privacy Act 1988 may apply when identifiers or behavioural data are collected. Clear consent notices, minimal collection and short retention periods turn that legal obligation into respectful visitor experience.
Equipment planning adds a further Australian consideration. Chargers, replacement cables and spare phones should be available, while any electrical or electronic products sold in Australia must meet applicable RCM and electrical safety requirements. Maintenance staff also need a short runbook that explains how to restart services, replace a failed unit and preserve the installation’s current state.
Make Synchronization Part of the Experience
Reliable timing should support the artwork rather than become a technical display of its own. Visitors perceive coherence when sound, projection, physical controls and mobile interfaces feel connected, even if each element reaches its response by a slightly different route. Designing several levels of response allows an installation to remain legible during minor delays and recover without presenting a blank screen or restarting the entire experience.
For collaborators, the strongest system is the one another person can diagnose. Versioned message definitions, documented device states, visible connection indicators and agreed recovery procedures make technical responsibility clearer. These practices also preserve artistic intent, because a team can change one part of the sequence without accidentally destabilising the rest.
Explore the ITP 30 Show online, experience its experimental projects, and support the artists and participants turning connected technology into meaningful public work.