Why this is the question that gets sized wrong
Most fleet deployments specify the in-vehicle hardware first (because it's the visible bit) and treat the headend as an afterthought. That works fine for the first 20-50 vehicles and then breaks. The headend SpeedFusion peer count caps out, throughput aggregates beyond the headend's hardware ceiling, or the licence model that seemed cheap at small scale becomes expensive at fleet scale. The fleet operator either upgrades the headend mid-rollout (expensive, operationally disruptive) or accepts performance degradation as the fleet grows.
The fix is to size the headend first, against the planned end-state fleet count, then specify the in-vehicle hardware to match. This post covers how.
The three sizing variables
Per-vehicle bonded throughput
What does each vehicle actually need to push? This sets the per-vehicle SpeedFusion bandwidth budget and constrains your in-vehicle hardware choice.
| Workload | Sustained throughput per vehicle | Burst |
|---|---|---|
| Telematics, GPS, basic CAD/dispatch | ~50 Kbps | 1-5 Mbps (incident upload) |
| Body-worn camera + telematics | 2-5 Mbps | 10-25 Mbps |
| Multi-camera passenger Wi-Fi | 10-25 Mbps | 50-100 Mbps |
| Live broadcast contribution (HEVC 1080p) | 5-12 Mbps sustained | 20+ Mbps |
| Live broadcast contribution (4K HEVC) | 25-40 Mbps sustained | 60+ Mbps |
| Heavy operational + multi-stream upload | 50-150 Mbps sustained | 200+ Mbps |
Critical point: it's the aggregate fleet throughput at peak that drives headend sizing, not the per-vehicle steady state. A 200-vehicle fleet of patrol units running 50 Kbps steady state with occasional 5 Mbps incident bursts has very different aggregate requirements from 200 vehicles streaming continuous body-worn video.
Total vehicle (peer) count
Each vehicle is a SpeedFusion peer at the headend. SpeedFusion peer entitlements are gated by the headend hardware and the SFN-LC licence tier:
| Headend hardware | Peers built in | Peers via SFN-LC licence |
|---|---|---|
| Balance 580X HW2 | 50 | Up to 300 |
| Balance SDX | ~150 | Up to thousands |
| Balance SDX Pro M2 | ~250 | Up to thousands |
| Balance 1350 EC | ~500 | Up to thousands |
| Balance 2500 EC | ~1000 | Tens of thousands |
| EPX-M8 | ~1000 | Tens of thousands |
| FusionHub Solo (cloud) | ~50-100 | Per-instance limits, scale by adding instances |
| FusionHub Pro (cloud) | ~500 | Per-instance limits, scale by adding instances |
The numbers are indicative; talk to us for specific platform/firmware combinations. The licence model matters: SFN-LC scaling has tier breakpoints that materially affect per-peer cost. Sizing for the right licence tier from day one is cheaper than upgrading mid-deployment.
Headend egress and aggregate throughput
The headend has to handle the aggregate fleet throughput on egress (typically to a private cloud, broadcast operations centre, or fleet management system). Headend egress is gated by:
- Router throughput at the headend hardware
- Internet/MPLS uplink at the headend
- Whatever sits beyond (fleet management system, recording infrastructure, etc.)
For most fleet deployments, the headend Internet uplink is the practical ceiling, not the router throughput. Plan accordingly.
The sizing framework
- Define the planned end-state fleet count. Not today's count. The maximum vehicle count you expect over the next 4-7 year hardware lifecycle.
- Calculate aggregate peak throughput. Per-vehicle sustained throughput, multiplied by the realistic concurrent peak (not all vehicles transmit simultaneously; budget for 60-80% concurrency depending on operations pattern).
- Pick the headend tier from the peer count and aggregate throughput tables above. Round up. Headroom is cheap; mid-deployment upgrades are expensive.
- Specify the SFN-LC licence tier matching peak peer count. Build in 30-50% headroom for fleet growth.
- Choose in-vehicle hardware from per-vehicle bonded throughput requirements and operating environment.
- Plan the headend Internet uplink for aggregate egress. If aggregate egress exceeds 1 Gbps, you're into multi-Gbps fibre territory.
- Validate against PrimeCare/EssentialCare+ pricing at fleet scale. Care plan cost across 500 vehicles is meaningful budget; size and quote it explicitly.
Worked example: 50-vehicle patrol fleet
Scenario
UK regional police force, 50 patrol vehicles, 15 traffic units, 5 unmarked. Body-worn camera uplink during operations, dispatch CAD telematics, ANPR plate-read telemetry, occasional incident video upload.
Per-vehicle throughput
Steady state: ~500 Kbps (telematics + light body-worn). Incident peak: 10-25 Mbps (active body-worn + ANPR alert + multi-camera). Concurrency at peak: ~30% (typical for patrol fleet of this size during major incident).
Aggregate at headend
Steady state: 50 vehicles × 500 Kbps = 25 Mbps sustained. Peak (30% concurrent at 25 Mbps each): ~375 Mbps. Plan for ~500 Mbps egress headroom.
Recommended headend
Balance 580X HW2 with optional FlexModule Mini 5G for cellular failover at the control room. 50 SpeedFusion peers built in (matches fleet count exactly; consider Balance SDX if growth is planned). PrimeCare J 1-Year through 4-Year for the headend; PrimeCare+ J for mission-critical 24/7 cover.
Recommended in-vehicle
Patrol vehicles: MAX BR1 Pro 5G for single-modem 5G or MAX BR2 Pro 5G for dual-cellular redundancy. Traffic units: MAX BR2 Pro 5G for the dual-cellular margin during pursuit operations. Mobility 42G antennas. PrimeCare C 4-Year per vehicle.
Worked example: 200-vehicle broadcast fleet
Scenario
National broadcaster operating 200 OB vehicles and ENG kits across UK. HEVC 1080p contribution from each unit during live transmission, occasional 4K UHD for premium events. Network operations centre headend in London.
Per-vehicle throughput
HEVC 1080p contribution sustained: 5-12 Mbps per vehicle. 4K UHD: 25-40 Mbps. Peak concurrency: ~25% of fleet active simultaneously (typical for national broadcaster operations).
Aggregate at headend
Steady state (50 vehicles active at 8 Mbps avg): 400 Mbps sustained. Peak (with 4K events and burst contribution): up to 1.5 Gbps. Plan for 2 Gbps egress headroom.
Recommended headend
Balance 1350 EC as the primary NOC SpeedFusion concentrator. 200 SpeedFusion peers comfortably within the 1350 EC's entitlement. SFN-LC licence tier matched to fleet count plus 30% headroom (260 peers). PrimeCare K 4-Year for the headend, with PrimeCare+ K for the mission-critical NOC operation.
Alternative for larger broadcasters
For broadcasters operating 500+ vehicles or planning rapid fleet expansion, the EPX-M8 with the LW1C3F3 pre-configured FlexModule selection (1x SIM Injector, 3x cellular, 3x fibre) is the broadcast NOC concentrator at scale.
Recommended in-vehicle
OB vehicles: MAX HD2 MBX 5G with Antenna MAX Duo mast-mounted. ENG kits: MAX Transit Pro E for single-5G or Transit Duo Pro for dual-cellular. EssentialCare+ at the appropriate tier per vehicle hardware.
Worked example: 500-vehicle delivery fleet
Scenario
UK national parcel delivery operator. 500 vehicles across rural and urban routes. Telematics, dispatch routing, route optimisation telemetry, occasional in-cab video upload for incident review. Control centre in Midlands.
Per-vehicle throughput
Steady state: ~200 Kbps (telematics + dispatch). Peak (incident review video upload): 5-10 Mbps. Concurrency at peak: ~10% (incident reviews are episodic).
Aggregate at headend
Steady state: 500 × 200 Kbps = 100 Mbps sustained. Peak (50 vehicles concurrent at 7 Mbps): ~350 Mbps. Plan for 500 Mbps egress headroom.
Recommended headend
Balance SDX as the headend with at least one FlexModule Plus for cellular failover. SFN-LC licence for 600+ peers. PrimeCare K 1-Year through 4-Year. Alternatively, Balance 1350 EC if the deployment plan includes future expansion beyond 1000 vehicles.
Alternative cloud-based headend
For deployments where the operator prefers a cloud-based concentrator (no on-premise router), Peplink FusionHub Solo or FusionHub Pro running on AWS, Azure, or GCP delivers equivalent SpeedFusion concentrator function. Pricing is by FusionHub instance + cloud compute. Talk to us for FusionHub vs hardware concentrator decision support.
Recommended in-vehicle
Bulk fleet: MAX BR2 Micro for compact LTE at moderate cost across 500 units. Higher-tier vehicles or routes with specific 5G requirements: BR1 Pro 5G or BR1 Mini 5G HW2. Future-proof IoT-tier deployments: BR1 Mini RedCap. Mobility 22G compact antennas. PrimeCare B 4-Year per vehicle.
Common sizing mistakes
Mistake 1: sizing for steady state, not peak concurrency
The headend has to handle peak. Sizing for steady-state aggregate looks great in the spec sheet and saturates during the first major incident. Always size for peak concurrency at peak per-vehicle throughput.
Mistake 2: under-licensing SFN-LC peers
Buying 100 SFN-LC licences for a 95-vehicle fleet leaves zero headroom for growth. The licence model has tier breakpoints; the right tier to specify is one above your end-state count.
Mistake 3: forgetting the headend Internet uplink
The headend can have all the throughput in the world; if the Internet uplink is 100 Mbps fibre, that's your aggregate ceiling. Spec the uplink for end-state aggregate plus 50% headroom.
Mistake 4: under-budgeting PrimeCare/EssentialCare+ at fleet scale
500 vehicles × PrimeCare C 4-Year (£556) = £278,000. That's a meaningful procurement line item that needs explicit budget. Size and quote it from day one.
Mistake 5: standardising the wrong in-vehicle tier
Specifying BR2 Pro 5G across an entire 500-vehicle fleet because some routes need dual-cellular over-spends on the routes that don't. Conversely, specifying BR1 Mini across the entire fleet under-spec'd for the 5% of vehicles that operate in coverage-marginal areas. Tier the fleet correctly: bulk on entry hardware, premium tier on the vehicles that need it.
Talk to us about fleet sizing
If you're scoping a fleet rollout and want sizing advice before you commit to architecture, talk to us. We size cellular antenna budgets against actual operational coverage data, not vendor specs. We design SpeedFusion peer architectures for end-state fleet count plus headroom. We quote PrimeCare/EssentialCare+ at fleet scale with multi-year breakpoints. And we'll tell you when the simpler architecture (smaller headend, fewer SFN-LC peers, more in-vehicle redundancy) is the better answer.
Email info@thetechfactory.co.uk with the deployment context (fleet count today, end-state plan, per-vehicle workload, target rollout window) or call +44 (0)1788 550000.
Further technical reading
For vertical-specific architectural detail, see our pages on marine networks, broadcast contribution, and public safety. For RedCap-specific fleet planning, see our 5G RedCap deep-dive. For vendor selection context, see Peplink vs Cradlepoint vs Fortinet.