Mixed Fleet: Human & Robot Warehouse
Almost no Australian warehouse goes from fully manual to fully autonomous in a single step. The realistic path runs through years of mixed-fleet operation, with human-operated forklifts and autonomous trucks sharing the same building. Done well, this is productive and safe. Done badly, it produces incidents, near-misses, and operator resentment that derails the whole automation business case. This page covers the operational design that makes mixed fleets work.
Three Mixed Fleet Models
Mixed fleets aren't a single thing. There are three distinct operational models, each with different design implications:
| Model | Description | Best For |
|---|---|---|
| Zone-segregated | Autonomous trucks in one half of the building, manual in the other; physical or virtual barriers between zones | Sites with stable task profiles by zone |
| Time-segregated | Autonomous trucks operate during off-peak hours; manual trucks during peak hours | Sites where peak demand needs human flexibility |
| Co-located | Both fleets operate in the same space simultaneously, coordinated by fleet management | Smaller buildings; tight integration |
Zone-Segregated Mixed Fleet (Recommended for Most)
The zone-segregated model is the most operationally robust starting point. Common zone splits:
- Bulk storage — autonomous (predictable, repetitive)
- Pick face — mixed (autonomous replenishment, manual picking)
- Receiving dock — manual or mixed (variable freight, exception handling)
- Despatch dock — autonomous (loading pre-staged trucks)
- Yard / outdoor — manual (weather variability, ad-hoc movements)
Zone boundaries should be enforced by both physical means (line-marking, low fences) and software (autonomous trucks treat zone boundaries as virtual walls). Zone transitions should be controlled gates with explicit handoff procedures.
Co-Located Mixed Fleet (Advanced)
Fleet Management Coordination
Co-located fleets require manual operators to interact with the fleet management system — either via tablet, dashboard, or in-cab terminal — so autonomous trucks know where manual trucks are operating.
Pedestrian Tracking
UWB or BLE tagging of pedestrians (operators, supervisors, pickers) gives autonomous trucks precise location awareness that LIDAR alone can't provide in cluttered environments.
Aisle Reservation Protocols
For narrow aisles where two trucks can't pass, fleet management implements aisle reservations — first-in claims the aisle, others queue. Manual operators see queue status on dashboard.
Manual Override Discipline
Manual operators must NOT enter zones marked autonomous-only without authorisation. Cultural discipline matters as much as technical controls; training and reinforcement are essential.
Operator Role Evolution
Mixed fleet operation reshapes the human operator role over time. The transition follows a typical pattern:
- Year 1: Concern — operators concerned about job security, watch autonomous trucks suspiciously, raise (often legitimate) safety concerns
- Year 2: Coexistence — operators settle into roles handling exception work, peak relief, and tasks autonomous trucks can't do well
- Year 3+: Specialisation — operators become fleet supervisors, troubleshooters, and high-skill manual operators handling the variable 20-30% of work that won't automate
Operator turnover is typically lowest in Year 1 (status quo bias) and highest in Year 2-3 as roles change. Plan for it.
Common Mixed Fleet Failure Modes
- Inadequate training — manual operators not trained on autonomous truck behaviour, leading to confused responses to autonomous truck movements
- Poor zone boundaries — fuzzy zone boundaries with autonomous trucks edging into manual zones (or vice versa)
- Mismatched pacing — autonomous trucks moving at fixed pace, manual operators moving at variable pace, leading to congestion and frustration
- Communication gaps — manual operator sees autonomous truck stopped, doesn't know if it's faulted or just paused, takes wrong corrective action
- WHS shortcuts — cultural normalisation of operating procedures degradation as the fleet matures
The Practical Recommendation
Start zone-segregated. Get 6-12 months of operational data. Then selectively expand autonomous coverage into adjacent zones based on actual operational data, not theoretical capability claims. Most operators find themselves at 60-75% autonomous coverage 24 months after the first deployment, with the remainder consciously kept manual for the work where human flexibility outperforms automation economics.