Schedule for Wednesday 27 May. 38 deliveries, 12 vehicles. 35 automatic, 3 escalations. MB4 remaining hours 1h45 but Genoa requires 2h20: option A 40-min earlier start, option B reassign to Clark (6h). Ford9 ADR for residential zone, ban 07-20: night window or intermediate hub.
AGENT · FLEET SCHEDULE OPTIMIZER
Logistics routes are optimized against the actual regulatory constraints.
Fleet Schedule Optimizer optimizes the customer's logistics fleet routes and shifts every day. It respects regulatory constraints: driving and rest times (Regulation EC 561/2006), port access rules, urban low-emission zones, ADR rules for hazardous goods transport. Cases with unresolvable conflicts escalate to the fleet manager.
Fleet Schedule Optimizer at work.
On MB4 reassign to Clark. On Ford9 intermediate hub at Sesto San Giovanni, night delivery.
Schedule updated. Shifts confirmed to drivers. Registry updated.
Why it exists.
Route optimization for a mid-to-large logistics operator involves hundreds of daily deliveries, dozens of vehicles, and multiple interlocking regulatory constraints. Transportation management systems (TMS) support operations management but have limits on dynamic multi-constraint optimization — particularly when remaining driving hours, low-emission zone restrictions, and port access rules all need evaluating together.
How it works each morning.
Fleet Schedule Optimizer activates every morning. It reads the scheduled deliveries, the available vehicles with their characteristics (load capacity, Euro class for low-emission zone access, ADR certification for hazardous goods), and the drivers' remaining driving hours under Regulation EC 561/2006. It produces the optimized schedule within the configured constraints. For standard pre-approved cases, scheduling proceeds automatically.
Conflicts go to the manager.
For cases with unresolvable conflicts or significant trade-offs — delivery close to the driving hours limit, zone with ADR restrictions, tight port access window — the agent escalates to the fleet manager with the available options.
The teams that manage fleet operational planning.
Fleet manager
Receives the optimized schedule every morning, with only the cases that genuinely require a human decision highlighted. Unresolvable conflicts arrive with concrete options, not as open problems to analyze from scratch.
Routes · 06:00 42 vehicles
Driving hours Reg. EC 561
Urban LTZ 2 detours
ADR · class 3 dedicated route
12 minutes saved per vehicle
Dispatcher
Stops doing multi-constraint optimization by hand. Their time focuses on exceptions and non-standard cases, not base scheduling.
Week 23 312 shipments
Madrid-South zone avg delay 28'
Complaint cluster 8 similar tickets
Carrier · Y performance ok
Brief to delivery lead
Transport compliance
Has the trace of each assignment with the regulatory constraints applied: driving hours respected, low-emission zone access verified, ADR rules respected. When the transport authority audits, the registry is queryable.
SAD · #SAD-2024-877 Reg. 952/2013
Goods code 8703.23
Origin DE · EU
ICS2 · pre-arrival transmitted
AEO record updated
An optimized daily schedule delivered before the shift starts.
Five-thirty: deliveries, drivers, vehicles, constraints.
For a regional carrier with twelve vehicles and forty daily deliveries, scheduling happens every morning at five thirty. The shift starts at seven. Fleet Schedule Optimizer reads the day's deliveries, the available drivers with their remaining driving hours calculated from the week's tachograph data, and the vehicle characteristics (two ADR-certified, three Euro 6 class for low-emission zone access, seven standard).
Thirty-five of forty automatic. Three escalations.
For thirty-five of the forty deliveries, scheduling is automatic. For three cases with conflicts — driver with insufficient remaining hours for the distance, residential zone with daytime ADR ban, tight port access window — the agent brings the fleet manager the concrete options with the timing and consequences of each.
The manager chooses. The drivers leave.
The fleet manager chooses in a few minutes on the work channel. The final schedule is sent to the drivers. Each assignment's data — constraint applied, conflict type handled, option chosen — enters the audit registry, readable by the transport compliance team.
Declarative rules from the customer's fleet operations team.
The rules of Fleet Schedule Optimizer are declarative. The customer's fleet operations and transport compliance team defines in a readable format the applicable regulatory constraints (driving/rest hours table by driver type, low-emission zone map with time windows, ADR-restricted zone list), the assignment priorities (criteria for choosing between equivalent options), and the escalation thresholds that require a human decision. The rules live in the customer's repository, versioned. TMS and digital tachograph integration is built during delivery via a dedicated adapter.
- Language
- TypeScript (Node.js)
- LLM model
- customer's choice: Anthropic, OpenAI, Mistral, open source models hosted internally, AWS Bedrock for a private model
- Built-in controls used
- pii-detector, topic-guardrail
- Native delivery channels
- Slack, Telegram, WhatsApp, OpenAI-compatible HTTP
- Scheduling
- configurable per instance (typical 05:00-06:00, before shift start)
- TMS integration
- dedicated adapter built during delivery by the Exelab team
- Digital tachograph system integration
- dedicated adapter built during delivery by the Exelab team
- Memory
- persistent per instance, pgvector + PostgreSQL FTS
- Registry
- immutable, queryable with a standard SQL client
How Fleet Schedule Optimizer works in detail.
No. Fleet Schedule Optimizer integrates with the existing TMS: it reads the scheduled deliveries, regulatory constraints, and driver data, produces the optimized schedule, and writes it back to the TMS. The TMS remains the fleet system of record. The agent adds the multi-constraint optimization layer that traditional TMS platforms handle with difficulty.
Constraint rules are declarative in the customer's repository. When a low-emission zone is updated, the transport compliance team modifies the declarative file, tests it in the development environment, promotes it to production. The rule update cycle stays entirely inside the customer's team, with no dependency on an external vendor for every change.
It depends on the configuration. Route optimization logic can include backhaul opportunities if the customer's TMS tracks them. The specific configuration is defined during discovery with the fleet operations team.
The typical pattern is 14-22 weeks. The range is wider than for other agents because it depends on TMS complexity, the variety of regulatory constraints to configure, and the number of operational exceptions to map. Discovery two weeks, rule configuration four weeks, TMS and tachograph integration four to six weeks, testing and hand-off to the fleet operations team two to three weeks.
From a 30-minute conversation to the squad in production.
A 30-45 minute conversation to understand how Fleet Schedule Optimizer would configure to the customer's case. TMS in use, priority regulatory constraints, fleet size and number of daily deliveries.