From Legacy Spreadsheets to Real‑Time Operations: A 12‑Week Logistics Control Tower for UrbanFleet

## Overview UrbanFleet is a mid‑size urban logistics operator serving same‑day and next‑day delivery across five Indian metros. They manage a mixed fleet of 380 vehicles (two‑wheelers, mini‑trucks, and vans), with third‑party subcontractors covering surge demand. As the business grew, their operating model did not: dispatchers used spreadsheets, drivers called in status updates, and customer support pulled data from four separate systems. The result was a reactive, fragile operation that could not scale. Webskyne was engaged to design and implement a real‑time operations control tower that consolidated fleet, order, and customer data, and introduced workflow automation for dispatch, exception handling, and performance reporting. The engagement ran for 12 weeks and included process redesign, data integration, a new operational dashboard, and role‑based tooling for dispatchers and supervisors. The solution needed to deliver measurable improvements quickly while remaining flexible for future growth. > **Cover image**: https://images.unsplash.com/photo-1469474968028-56623f02e42e?auto=format&fit=crop&w=1600&q=80 --- ## Challenge UrbanFleet’s operational issues were not driven by a single broken system; rather, they were the result of fragmented information and manual processes. The core challenges included: 1. **No single source of truth**: Order status lived in the order management system, vehicle locations in a telematics dashboard, and customer support tickets in a CRM. Dispatchers manually reconciled these in Excel each morning and through the day. 2. **Reactive exception handling**: Delays were discovered only after customers called. There was no proactive alerting or prediction. 3. **Low utilization**: Subcontracted vehicles were under‑used because route assignment happened late, and fleet managers lacked visibility into availability. 4. **Inconsistent SLAs**: Operations could not consistently hit promised delivery windows. During peak days, late delivery rates rose above 18%. 5. **High support load**: Customer service spent 30–40% of their day responding to “Where is my order?” calls. 6. **Data silos and misaligned KPIs**: Operations, sales, and customer support each tracked different KPIs, leading to unclear priorities. The business risk was significant: UrbanFleet’s largest enterprise client had a renewal decision in 90 days and had already raised concerns about reliability. --- ## Goals The engagement defined measurable goals across performance, visibility, and process adoption: - **Improve on‑time delivery rate by at least 20%** within 60 days of launch. - **Reduce customer support “Where is my order?” tickets by 30%** through proactive updates. - **Increase vehicle utilization by 15%** via better dispatching and workload balancing. - **Create a real‑time operational view** that consolidates order, fleet, and support data into a single dashboard. - **Enable proactive exception management** with automated alerts and recommended actions. - **Deliver within 12 weeks**, with minimal disruption to ongoing operations. The solution also needed to be extensible: UrbanFleet planned to add two new cities in the next quarter and required an architecture that could scale with them. --- ## Approach Webskyne used a “control‑tower first” strategy: establish a unified operational view, then automate exceptions and workflows around it. The approach was broken into five phases: 1. **Discovery & process mapping** (Weeks 1–2): Shadowed dispatch and customer support teams, mapped the current order lifecycle, and cataloged all data sources. 2. **Data integration foundation** (Weeks 3–5): Built ingestion pipelines for telematics, OMS, and CRM data. 3. **Control tower MVP** (Weeks 6–8): Delivered the first real‑time dashboard with live tracking and predictive ETAs. 4. **Workflow automation** (Weeks 9–10): Added alerting, exception queues, and SLA rules. 5. **Adoption & performance tuning** (Weeks 11–12): Trained teams, tuned scoring models, and deployed performance reporting. Each phase shipped incrementally, ensuring stakeholders saw value early and could shape the direction as the system evolved. --- ## Implementation ### 1) Data architecture & integration UrbanFleet had three core systems plus manual spreadsheets. The data strategy prioritized near‑real‑time visibility without disrupting existing infrastructure. We implemented a lightweight event pipeline and adopted a “read‑only” integration pattern to reduce risk. **Key components:** - **Telegest API ingestion**: Vehicle location pings and status updates were pulled every 60 seconds. - **OMS integration**: Order lifecycle events were streamed via a webhook proxy. - **CRM sync**: Support tickets were fetched every 5 minutes and enriched with order metadata. - **Master data store**: A PostgreSQL instance served as the control‑tower backbone, with tables for orders, vehicles, hubs, and drivers. - **Event processing**: A NestJS worker processed events and generated derived metrics such as ETA confidence and delay risk. To ensure reliability, the pipeline supported backfills, replay, and idempotent processing. The architecture was intentionally modular to allow UrbanFleet’s team to extend it later without vendor lock‑in. ### 2) Control tower dashboard We designed the control tower to match the mental model of dispatchers and supervisors. The dashboard emphasized visibility and rapid decision‑making, not analytics for its own sake. **Key views included:** - **Live fleet map** with route overlays and delivery status coloring. - **Exception queue** with severity scoring (e.g., at‑risk, delayed, stuck). - **Hub performance panel** with cycle time, idle time, and capacity utilization. - **SLA heatmap** for each customer segment. The UI was built in Next.js with real‑time updates via WebSockets. We kept the interface minimal and used bold status indicators to reduce cognitive load. > **In‑line image**: https://images.unsplash.com/photo-1489515217757-5fd1be406fef?auto=format&fit=crop&w=1400&q=80 ### 3) Predictive ETA scoring UrbanFleet needed to shift from “reactive updates” to “proactive alerting.” We introduced a lightweight ETA scoring model using historical delivery data and real‑time signals. **Signals used:** - Vehicle location and speed - Route congestion indicators - Historical lane performance (hub‑to‑hub timings) - Driver reliability scores - Time‑of‑day and day‑of‑week patterns The model assigned each active order an ETA confidence score. Orders below a defined threshold entered the exception queue automatically, and dispatchers could trigger pre‑emptive actions such as re‑routing or customer notification. ### 4) Workflow automation & alerting We implemented automated workflows with clear escalation paths: - **Delay alerts** triggered at 20% and 40% ETA risk thresholds. - **Idle vehicle alerts** triggered when a vehicle remained inactive for 30 minutes after hub assignment. - **Customer proactive updates** sent to the support team’s CRM with suggested scripts. - **Auto‑reassignment** recommendations for low‑priority orders in overload scenarios. These workflows were tuned collaboratively with dispatch managers to ensure they aligned with actual operating constraints. ### 5) Change management and training Technology alone does not change operations. We ran structured training sessions for 30 dispatchers and 8 supervisors, created a quick‑start playbook, and embedded “office hours” during the first two weeks post‑launch. We also aligned KPIs across operations and customer support so that all teams were working toward the same outcomes. --- ## Results The control tower went live in Week 10 with two pilot cities, followed by a rapid rollout to the remaining three metros in Week 12. Within 60 days, UrbanFleet reported measurable improvements across its key metrics. ### Performance outcomes - **On‑time delivery rate improved by 24%** (from 76% to 94%). - **Late deliveries reduced by 41%** across all routes. - **Vehicle utilization increased by 17%**, driven by earlier assignment and real‑time availability. - **“Where is my order?” tickets dropped by 38%**, freeing support capacity. - **Average dispatch response time decreased by 52%**, as exceptions were visible in a single queue. ### Operational impacts - Dispatch managers spent **30% less time reconciling spreadsheets**. - Supervisors gained a real‑time view of hub health and could rebalance fleet resources during surges. - The enterprise client renewed for another year, citing “significant improvement in delivery predictability.” ### Financial impact (estimated) UrbanFleet estimated that improved utilization and fewer SLA penalties reduced monthly operational costs by **₹18–22 lakh**, with the system expected to pay back its investment within two quarters. --- ## Metrics Snapshot (60‑Day Post‑Launch) - **On‑time delivery:** 76% → 94% - **Late deliveries per 1,000 orders:** 180 → 106 - **Average idle time per vehicle/day:** 68 min → 46 min - **Support tickets (WISMO):** 1,200/month → 740/month - **Dispatch escalations:** 85/week → 41/week - **Enterprise SLA compliance:** 82% → 96% --- ## Lessons Learned ### 1) Visibility beats complexity UrbanFleet didn’t need a massive analytics platform; they needed a clear, live operational view. By focusing on real‑time visibility first, adoption was faster and the system immediately delivered value. ### 2) Align KPIs early Operations and customer support initially had conflicting goals. Bringing them together around shared SLA metrics reduced friction and improved follow‑through when exceptions occurred. ### 3) Automation must match reality We learned that automated recommendations only work when aligned with on‑ground capacity. Early versions of the auto‑reassignment feature were too aggressive and ignored local hub constraints. After tuning thresholds and embedding supervisor approval, the workflow became trusted and effective. ### 4) Incremental rollout reduces risk Launching in two pilot cities provided space to tune the ETA model and train teams without overwhelming the organization. This phased rollout increased confidence and built momentum. ### 5) Change management is a product The training playbook, in‑app tips, and daily office hours were just as critical as the technology. The tools were adopted because people understood the “why” and saw their daily workload improve. --- ## Final Takeaway For logistics companies, operational reliability is not just about vehicles and routes—it’s about real‑time decisions. By consolidating data, surfacing exceptions early, and aligning teams around shared metrics, UrbanFleet transformed its operations in 12 weeks. The control tower now acts as the single source of truth, empowering dispatchers and supervisors to make proactive decisions that improve delivery performance and customer satisfaction. At Webskyne, this case study reinforced our approach: start with visibility, build workflows around real operational constraints, and prioritize adoption as much as the technology. The results show that when operations become data‑driven in real time, the impact is both measurable and immediate.