VYUH SUPPLY CHAIN NEURAL NETWORK · AGENT 05

TOM

LOGISTICS AGENT
The typhoon hit at 2am.
Tom rerouted 14 shipments by 2:14am.

Tom monitors every freight lane, every vessel, every weather system, and every geopolitical risk — simultaneously, continuously. It sees disruptions 72 hours before they happen and reroutes autonomously in 14 minutes. No emergency calls. No chaos. Just delivery.

340
Lanes Monitored
Every freight route, sea, air, road — globally
72hr
Disruption Warning
Average warning before a lane failure hits your cargo
14min
Crisis Response
From disruption detection to rerouted + carrier booked
£6M
Emergency Freight Saved
Per major disruption vs reactive routing
What Tom Does

Six ways Tom keeps your goods moving

Tom doesn't react to disruptions. It predicts them, routes around them, and keeps Becci's production schedule intact.

🌊
Continuous lane risk monitoring
Tom monitors all 340 active freight lanes 24/7 — ocean, air, road, rail. AIS vessel tracking, port congestion data, weather forecasts, and geopolitical risk feeds synthesised into a real-time lane risk score for every route carrying your cargo.
340 LANES · CONTINUOUS · ALL MODES
72-hour disruption forecasting
Tom's ensemble prediction model identifies lane disruptions an average of 72 hours before they become critical — combining meteorological models, port operations data, carrier network feeds, and political risk intelligence into a probabilistic disruption timeline.
72HR FORECAST · ENSEMBLE MODEL
🔄
Autonomous rerouting
When disruption probability exceeds the threshold, Tom solves the rerouting problem immediately — calculating alternative lanes, comparing transit times and costs, selecting the optimal carrier, and raising the booking autonomously. All within 14 minutes.
14MIN RESPONSE · CARRIER BOOKED
💰
Freight cost optimisation
Tom continuously re-optimises freight routing against cost — consolidating shipments where possible, timing dispatches to avoid peak surcharges, selecting carriers against SLA performance history, and capturing rate arbitrage across modes. Each optimisation decision logged and reportable.
CONTINUOUS · MULTI-MODAL
📊
Carrier SLA tracking
Every carrier, every lane, every shipment — Tom tracks on-time performance, damage rates, documentation accuracy, and claims resolution speed. Carriers with declining SLA performance are flagged and progressively de-weighted in the routing algorithm before they cause a failure.
PER CARRIER · PER LANE · REAL-TIME
🏭
Production schedule protection
Tom receives Becci's production schedule and works backwards to identify which inbound shipments are on the critical path. These shipments receive elevated monitoring, faster escalation thresholds, and automatic alerts to Becci and Cho if timeline risk emerges.
COORDINATES WITH BECCI + CHO
Where Tom Gathers Information

Six external feeds and two agent signals

Tom fuses satellite, meteorological, geopolitical, and commercial data into a single real-time risk picture across every lane.

🛸
AIS Vessel Tracking
MarineTraffic AIS data — live position, speed, heading, and ETA for every vessel carrying client cargo. Anomalies in vessel behaviour trigger immediate risk flags.
LIVE · MARINÉTRAFFIC AIS
🌩️
Satellite Weather Feeds
NOAA and ECMWF meteorological models — typhoon tracks, storm systems, extreme weather events — processed against active freight lanes to calculate weather disruption probability.
NOAA · ECMWF · HOURLY
Port Congestion Data
Real-time port authority feeds — vessel queue lengths, berth availability, customs clearance times, industrial action notices, and terminal closures across all major global ports.
PORT AUTHORITY · GLOBAL COVERAGE
🚢
Carrier APIs
Direct API connections to carriers — Maersk, MSC, CMA CGM, DHL, FedEx, and others — for live vessel position, booking availability, rate quotes, and confirmed booking status.
DIRECT CARRIER APIS · BOOKING CAPABLE
🌍
Geopolitical & Labour Risk
Political risk intelligence feeds, trade restriction databases, sanctions lists, and labour action monitoring — covering port strikes, customs policy changes, and trade route restrictions across 180 countries.
POLITICAL RISK · 180 COUNTRIES
📈
Freight Rate Markets
Spot rate indices (Baltic Dry, SCFI, Freightos) and contract rate data — allowing Tom to identify rate arbitrage opportunities and optimise freight cost against market conditions.
BALTIC DRY · SCFI · FREIGHTOS
🏭
Becci Production Schedule
Tom receives Becci's published production schedule and identifies which inbound materials are on the critical path. Critical path shipments receive heightened monitoring and faster escalation.
AGENT SIGNAL · ON EVERY SCHEDULE
🔄
Cho Rebalancing Instructions
When Cho detects inter-warehouse inventory imbalance, Tom receives the rebalancing instruction and coordinates the inter-site freight — carrier selection, booking, and tracking.
AGENT SIGNAL · ON REBALANCE
ENSEMBLE DISRUPTION PREDICTION + DYNAMIC VEHICLE ROUTING (DVRP)
Tom's disruption prediction model is an ensemble combining a gradient-boosted classifier for lane risk scoring, an LSTM sequence model for temporal pattern detection in weather and port data, and a Bayesian network for geopolitical risk propagation. When disruption probability exceeds threshold, the rerouting problem is formulated as a Dynamic Vehicle Routing Problem (DVRP) and solved using a combination of Clarke-Wright savings algorithm and tabu search — delivering a cost-optimal alternative routing within the 14-minute response window.
How Tom Thinks

Five-step logistics intelligence cycle

01
📡
Continuous monitoring
All 340 lanes monitored continuously. AIS, weather, port, carrier, and geopolitical feeds ingested and fused into a lane risk score per active shipment. Real-time — not a daily batch.
AUTOMATIC
02
🔮
Ensemble disruption prediction
Gradient-boosted classifier + LSTM sequence model + Bayesian geopolitical network. When any lane risk score exceeds threshold, rerouting is triggered immediately.
AUTOMATIC
03
🗺️
DVRP rerouting solver
Dynamic Vehicle Routing Problem solved in real time — evaluating alternative lanes, transit times, carrier availability, and cost. Optimal reroute identified from 340 lane options within 14 minutes.
AUTOMATIC
04
🚦
Confidence gate
Reroute solutions above 82% confidence execute autonomously — carrier booked, Becci and Cho notified. Between 65–82%, a logistics team review flag is raised. Below 65%, a full escalation briefing is generated.
AUTOMATIC
05
📤
Execute & notify
Carrier booking raised. Becci notified of any production timeline impact. Cho updated on inter-warehouse movements. Commercial team briefed on cost delta vs original routing plan.
AUTOMATIC
What Tom Produces

Six outputs protecting your supply lines

🗺️
Lane Risk Assessment
Daily risk report for all active freight lanes — risk score, primary risk driver (weather / geopolitical / congestion / carrier), and recommended monitoring level. Board-ready summary with financial exposure calculation.
DAILY · TO LOGISTICS + COMMERCIAL
Taiwan Strait: RISK 84/100 · Typhoon Kai ETA 18hr · 14 shipments at risk · £6.2M exposure
Disruption Alert
72-hour ahead alert when Tom's ensemble model flags a lane disruption probability above threshold. Includes disruption type, ETA, affected shipments, and preliminary rerouting options with cost comparison.
REAL-TIME · TO LOGISTICS TEAM + BECCI
ALERT: Taiwan Strait · Typhoon · 14 shipments · Options: Singapore reroute +2d £1.2M | Air freight +£5.1M
🔄
Rerouting Instruction + Booking
When Tom executes a reroute autonomously, a structured booking is raised with the new carrier, confirmed transit time, revised delivery date, and cost delta. All documentation updated automatically.
AS REQUIRED · CARRIER BOOKING CONFIRMED
REROUTE EXECUTED: 14 shipments via Singapore · Carrier: Maersk · ETA: on schedule · Cost delta: +£1.2M
💰
Freight Cost Optimisation Report
Monthly report showing freight cost per lane vs benchmark, consolidation opportunities identified and captured, carrier rate arbitrage captured, and total savings versus reactive procurement baseline.
MONTHLY · TO PROCUREMENT + FINANCE
Month saving: £840K · Consolidations: 18 · Rate arbitrage captured: 8 lanes · Emergency avoided: 3
📊
Carrier SLA Performance Report
Weekly report showing on-time delivery performance, damage claims, documentation accuracy, and claims resolution speed by carrier and lane. Declining performers flagged before they cause a live failure.
WEEKLY · TO PROCUREMENT + LOGISTICS
WATCH: Carrier X · On-time: 71% (was 94%) · Lane: Shanghai→Rotterdam · Weighting reduced automatically
🏭
Production Delay Notification
When a reroute extends a critical-path delivery beyond Becci's latest acceptable date, Tom notifies Becci immediately with revised ETA and material arrival date — allowing proactive production replanning rather than a forced emergency.
AS REQUIRED · TO BECCI + PLANNING
NOTIFY BECCI: PCIe connectors · Original ETA: 12 Apr · Revised: 14 Apr · 2-day buffer remains
Impact

What Tom changes in the first 90 days

£6M
Saved per disruption event
Emergency freight costs avoided by rerouting 72 hours early vs reacting after the event.
72hr
Average advance warning
From disruption detection to rerouting decision — before the crisis reaches your cargo.
14min
Crisis to reroute complete
Carrier booked, Becci and Cho notified — all within 14 minutes of the disruption being confirmed.
23%
Freight cost reduction
Through continuous route optimisation, consolidation, and rate arbitrage across 340 lanes.
See the Simulation

Watch Tom reroute 14 shipments in 14 minutes

A pre-recorded simulation of Tom detecting Typhoon Kai approaching the Taiwan Strait, calculating risk to 14 active shipments, and executing a full reroute through Singapore — all in 14 minutes, at 2am, without a single human decision. This is what your supply chain looks like.

🌊
Tom · Disruption Response Simulation
Typhoon Kai · Taiwan Strait · 14 shipments at risk · £6.2M exposure
SCENARIO
Typhoon Kai · Taiwan Strait · Cat 3 · ETA 18 hours
DETECTION TIME
02:00:01 local time
SHIPMENTS AT RISK
14 vessels · £6.2M cargo value
REROUTE COMPLETED
02:14:33 — 14 minutes
▸ TOM REROUTE EXECUTION — TYPHOON KAI RESPONSE
Disruption typeTyphoon Cat 3 · Taiwan Strait · 18hr ETA
Shipments affected14 vessels · 340 TEU · £6.2M value
Original routeTaiwan → Shanghai → Rotterdam
Reroute selectedTaiwan → Singapore → Rotterdam (+2 days)
Carrier bookedMaersk · 14 confirmed slots · departure 06:00
Emergency freight cost saved£6M vs reactive scenario
Becci notified2-day buffer remains — no production impact
Total response time14 minutes 32 seconds — autonomous
Situation Assessment
At 02:00:01, Tom detected Typhoon Kai entering the Taiwan Strait — Category 3, wind speeds 185 km/h, estimated arrival 18 hours. 14 active shipments carrying 340 TEU of client cargo are on the affected route — combined value £6.2M. NOAA trajectory models give a 94% probability of the typhoon reaching the strait within the 18-hour window. The Taiwan Strait route is being closed to commercial traffic. Tom immediately began calculating the optimal rerouting solution.
Actions Taken
01
All 14 shipments rerouted via Singapore. Tom evaluated 6 alternative routes across 340 available lanes. Singapore via Malacca Strait selected as optimal — adds 2 days transit time, £1.2M additional cost versus £7.2M emergency freight premium if rerouted reactively after the typhoon.
02
Maersk booking confirmed. 14 vessel slots confirmed for 06:00 departure. Carrier confirmed identical documentation and customs clearance handling. All Bills of Lading amended automatically.
03
Becci notified. Production schedule reviewed against revised ETAs — 2-day buffer remains on all critical-path materials. No production impact. Becci confirmed no replanning required. Cho updated on revised inventory arrival timing.
Financial Impact
£6M
Emergency cost avoided
+£1.2M
Reroute premium (vs £7.2M reactive)
14 min
Total response time
Recommendation
All 14 shipments are on the Singapore reroute, departing 06:00 today. No production impact has been identified — all critical-path materials have 2-day buffer. Tom is continuing to monitor Typhoon Kai's track for any further deviation. If the typhoon dissipates before full track recovery, Tom will evaluate reverting to original routing on next cycle. No escalation required.
"This is what your supply chain looks like under disruption. Book a session and we run it live on your freight lanes."
📅 Book my simulation session →
Book a Session

See Tom run on your freight lanes

Request a simulation session
Tell us about your logistics operation and we will come prepared with a simulation calibrated to your freight lanes, carrier mix, and key disruption scenarios.
Session requested
We will be in touch to confirm a time. Tom will be ready for your freight lanes.
Talk to Tom

Ask Tom about your logistics

Configure your freight operation below and every response becomes specific to your lanes, carriers, and disruption scenarios.

🌊
TOM
Logistics Agent · Monitoring 340 lanes
✓ Configured
🌊
TOM
I monitor every freight lane carrying your cargo — 24/7, across all modes. I can tell you which lanes are at risk right now, how much your reactive routing is costing you versus proactive rerouting, and which of your carriers are quietly degrading. Configure your operation above for specific analysis, or ask me anything.
VYUH Technical White Paper — Tom
Predictive Disruption Management in Global Freight Networks: The Tom Autonomous Logistics Agent
VYUH Supply Chain Neural Network · Technical Research Series · 2026
Logistics Systems Architecture · Applied Operational Research
ABSTRACT

This paper presents Tom, the autonomous logistics agent within the VYUH Supply Chain Neural Network. Tom employs an ensemble disruption prediction model — combining gradient-boosted classification, LSTM sequence modelling, and Bayesian geopolitical risk propagation — to forecast freight lane disruptions an average of 72 hours before they become critical. Upon disruption detection, Tom solves the rerouting problem as a Dynamic Vehicle Routing Problem (DVRP) and executes the optimal reroute autonomously within 14 minutes. We describe the full technical architecture, prediction model design, DVRP formulation, confidence gating mechanism, and integration within the VYUH agent network.

FREIGHT NETWORK — DISRUPTION REROUTE MAP TAIWAN SINGAPORE SHANGHAI BYPASSED ROTTERDAM DUBAI TAIWAN STRAIT TYPHOON KAI · CAT 3 ETA 18hr · 14 shipments at risk REROUTE COMPLETED 14 vessels · Singapore route Response time: 14 minutes £6M emergency cost avoided Original (disrupted) Active reroute

1. Introduction

Global freight networks are inherently exposed to disruption — typhoons, port strikes, geopolitical restrictions, vessel failures, and carrier capacity constraints cause regular service interruptions across every major trade lane. The economic cost of these disruptions falls disproportionately on companies that react to disruptions rather than anticipating them. Emergency freight premiums, expediting costs, and production downtime caused by delayed materials represent a significant and largely avoidable cost category.

The fundamental problem is one of information and response speed. Disruptions are often predictable days in advance from publicly available data sources — weather models, port operations feeds, AIS vessel tracking — but the integration and interpretation of these signals across hundreds of active freight lanes exceeds human cognitive bandwidth. By the time a logistics planner identifies a developing disruption and begins evaluating rerouting options, the low-cost rerouting window has often closed.

Tom addresses this through continuous, automated monitoring and instant autonomous response — detecting disruptions from multi-source signals, solving the rerouting problem mathematically, and executing the optimal solution before the disruption window closes.

2. Architecture

Tom's architecture comprises four layers: (1) a multi-source data ingestion layer processing AIS vessel feeds, NOAA/ECMWF weather models, port authority data, carrier APIs, and geopolitical risk feeds in real time; (2) an ensemble disruption prediction model that synthesises these inputs into per-lane risk scores with 72-hour forecasting horizon; (3) a Dynamic Vehicle Routing Problem (DVRP) solver that calculates optimal rerouting solutions when disruption is confirmed; (4) a confidence-gated execution layer that books carriers and notifies inter-agent peers autonomously for high-confidence solutions.

3. Ensemble Disruption Prediction Model

Tom's disruption prediction model is an ensemble of three components. The primary component is a gradient-boosted classifier (XGBoost) trained on a 5-year historical dataset of lane disruptions paired with pre-disruption signal features — including weather model outputs, port congestion metrics, AIS anomaly patterns, and geopolitical risk indices. This model achieves 89% precision and 84% recall at the 72-hour horizon for major disruptions.

The second component is an LSTM sequence model that processes time-series signals — particularly AIS vessel behaviour patterns and port throughput trends — to identify temporal anomalies that precede lane disruptions but are not captured by static feature models. The LSTM adds approximately 8 percentage points of recall for disruptions with gradual onset (port congestion, slow-developing weather systems).

The third component is a Bayesian network for geopolitical risk propagation — modelling how political events in one region propagate through connected shipping routes and port operations. This is particularly relevant for trade restriction events and port industrial action, which spread through the network in predictable patterns.

The three model outputs are combined through a learned ensemble weighting that is recalibrated quarterly against recent prediction performance. The final disruption probability estimate triggers the DVRP solver when it exceeds a lane-specific threshold calibrated to the cost of unnecessary rerouting versus the cost of a missed disruption.

4. Dynamic Vehicle Routing Problem Formulation

When disruption probability exceeds threshold, Tom formulates the rerouting problem as a DVRP. The problem is defined over a graph G = (N, E) where nodes represent ports and intermediate waypoints, edges represent freight lanes with associated cost, time, and capacity attributes. The objective is to find the minimum cost routing for all affected shipments that satisfies delivery time constraints and carrier capacity constraints.

The DVRP is solved using a two-phase approach: an initial solution is generated using the Clarke-Wright savings algorithm, then improved using a tabu search metaheuristic with a 10-minute computation budget. This consistently produces near-optimal solutions — within 3% of the mathematical optimum — while meeting the 14-minute total response time target (which includes data ingestion, problem formulation, solving, and carrier booking).

5. Confidence Gating and Autonomous Execution

Rerouting solutions are subject to a composite confidence gate before autonomous execution. The confidence score integrates: disruption model certainty (how confident is the ensemble model in the disruption prediction?), DVRP solution quality (how close is the solution to the theoretical optimum?), carrier booking confirmation (has the alternative carrier confirmed slot availability?), and production schedule impact assessment (has Becci confirmed the revised timeline is acceptable?). Solutions above 82% confidence execute autonomously. Between 65% and 82%, a logistics team review flag is raised with a 30-minute window. Below 65%, a full escalation briefing is generated.

6. Inter-Agent Integration

Tom's integration with the broader VYUH agent network is essential to its production-protecting function. Becci's production schedule identifies the critical path materials — the specific inbound shipments whose delay would cause a production halt. Tom applies heightened monitoring to these shipments and lower disruption thresholds, ensuring that critical materials receive the earliest possible warning and the most aggressive response. When a reroute modifies a critical-path delivery timeline, Tom notifies Becci immediately with a revised ETA — triggering proactive replanning rather than an emergency response. Cho receives revised inventory arrival timing, updating the stockout risk model accordingly.

7. Results

Over 12 months of operation across a 340-lane global freight network: average disruption warning time 72 hours (vs industry standard of same-day reactive response). Average reroute response time 14 minutes. Emergency freight cost savings: £6M per major disruption event. Freight cost reduction through continuous route optimisation: 23%. Carrier SLA degradation caught pre-failure: 94% of cases (vs 31% in manual monitoring environment). Production delays caused by logistics failures: reduced by 87%.

8. Limitations

Tom's disruption prediction model has lower accuracy for novel event types with no historical precedent in the training data. The 72-hour warning horizon reflects average performance — some disruption types (sudden geopolitical events, unexpected port industrial action) occur with shorter warning windows, and Tom's response in these cases is reactive rather than anticipatory. The DVRP solver assumes carrier slot availability can be confirmed rapidly; in periods of extreme industry-wide disruption, carrier capacity constraints may extend the booking confirmation step beyond the 14-minute target. These limitations are disclosed transparently in every disruption response briefing generated by Tom.

VYUH Supply Chain Neural Network · Tom · Logistics Agent
Sara Jr Ari Becci Cho