Segway Loomo представив автономного робота-доставника на CES 2019

Recommendation: Start a controlled pilot of Loomo for last-mile tasks on a campus or corporate site, with a small fleet of 3–5 units and a waitlist of potential customers to validate reliability and package handoffs before scaling.

At CES 2019, Segway Loomo debuted as a driverless delivery robot, highlighting спритність as it navigates sidewalks and curbside spaces. The booth emphasized scooters-style mobility paired with a robotic chassis, and a short movie-style clip circulated, noting Loomo’s potential to power hundreds of curbside deliveries and everything from obstacle avoidance to smooth handoffs, with demonstrations expected to scale in the coming years, especially where operators manage last-mile flows.

The show-floor course spanned hundreds of meters, and operators monitored the run while Luke narrated a short movie-style clip. Loomo combines SLAM-based mapping with a driverless control loop to negotiate pedestrians, curbs, and temporary barriers, showing спритність in real-world edge cases and a pace suitable for last-mile tasks.

For operators weighing a transition to autonomous delivery, Loomo’s CES 2019 showing suggests a practical path: begin with driverless trials in controlled environments, then scale to a fleet на right-sized routes. The data from pilots can inform cost-per-delivery and reliability, helping you move from speculated outcomes to evidence over the next years. Given the buzz around automation, many speculated outcomes exist, but concrete pilots tied to a defined route and a robust safety framework will determine whether Loomo-like units become a standard part of last-mile operations, bridging the chasm between manual and autonomous work.

CES 2019 Loomo Delivery: Practical implications for sidewalk robotics and last-mile logistics

Start with a city-scale pilot in a mid-size city using a small fleet of sidewalk delivery devices to handle parcels along fixed routes, with remote operators ready to intervene and a centralized fleet-management system to maintain safety and reliability. Each device runs the same software profile to simplify training and maintenance. This approach also focuses on innovating last-mile logistics while building public confidence and a repeatable playbook for other operators around the country.

Standardize hardware and software so a single device profile can support various routes and city blocks; the device relies on wheels for stability on sidewalks and sits on a four-point base to absorb shocks. Implement geofencing, audible warnings, and pedestrian-friendly speeds to protect walkers; most handoffs occur near storefronts or curb edges, and the system can have more than a handful of daily cycles on a single charge.

Capacity and throughput show practical implications: a small transporter fleet can move hundreds of parcels per day; over years, the supply chain gains scale to a million packages, easing last-mile pressure for city retailers and courier networks.

Public reception hinges on cultural expectations; novelty fades as routines prove reliable. The edge of adoption depends on comfort with on-street deliveries that are quiet, predictable, and respectful of pedestrians; around parks and transit hubs, residents notice the devices less and rely on predictable service. For those who enjoy bird-watching or simple commutes, the presence becomes normal.

Tech partnerships enable smoother operations: elevenlabs provides clear voice confirmations to customers; wayve contributes to perception and routing; james told attendees that operators focus on edge cases, which keeps service reliable and routes continually improve through real-world data and applications, given these factors. Start with high-density routes, measure customer satisfaction and delivery times, and expand to adjacent neighborhoods.

Loomo’s hardware and sensor suite: enabling precise sidewalk deliveries

Use Loomo’s self-balancing platform with a layered sensor stack and a robust local SLAM to achieve precise sidewalk deliveries. Elad says this pairing delivers centimeter-grade localization about similar pavements while maintaining smooth operation around pedestrians. Noting the past tests by startups in pilots, the combination gives operators a reliable baseline for last-mile tasks, while this setup stays adaptable to changing routes and conditions.

The hardware and sensor suite includes a stereo camera pair, a compact LiDAR module, ultrasonic sensors for close-range detection, a high-precision IMU, wheel encoders for odometry, and a GPS/RTK module for map-relative positioning. The onboard compute runs a fusion pipeline that integrates vision, depth data, and inertial measurements, delivering reliable інформація for planning and safe navigation. Payload capacity reaches up to 5 kg, enabling common last-mile parcels without sacrificing balance or range. The kit unveiled at CES 2019 demonstrates the hardware-software integration at scale.

For calibration and maintenance, Loomo’s perception stack supports quick re-calibration after payload changes and routine system checks to maintain sensor alignment. The fusion pipeline maintains low latency, allowing the robot to reroute quickly when obstacles appear, while redundancy across sensors helps it handle weather- or lighting-induced noise. Most deployments rely on GPS/RTK for corridor alignment and LiDAR for obstacle avoidance, with the vision system filling gaps during GPS outages.

When deploying this model, operators should partner with local logistics teams and city partners to align routes, curb-crossing rules, and pedestrian safety practices. The system’s modular design lets startups and established company teams swap sensors or add compute as the market evolves, keeping everything compatible with existing fleets. Elad notes that this openness supports a steady pipeline of deployments with likeminded partners, while information gathered from each run informs updates to the product roadmap. To maintain stability, operators should avoid rash feature changes and roll out updates in controlled pilots.

Practical recommendations for operators: pre-map sidewalks with high-quality data, test in similar environments, and document sensor performance across lighting and weather. Use data privacy controls and logging to protect information while supporting iterative improvement. With a scalable model and a clear path to expansion, the hardware and sensor suite can support million deliveries across multiple markets over time.

Navigation and obstacle avoidance: how Loomo manages pedestrians and curbs

Enable pedestrian-safe mode with a 1.5-meter buffer and cap speed at 4–5 km/h in crowded areas to reduce late braking and keep pedestrians safe.

Loomo’s navigation relies on a sensor fusion stack that blends stereo cameras, ultrasonic sensors, and IMU data, with SLAM-based maps to locate curbs and track pedestrians in real time, noting traffic dynamics and street layout so everything stays within a safe corridor.

When a walker enters the path, Loomo predicts trajectory over the next few seconds and re-routes to a safer line; a curb edge detection triggers an immediate stop and re-planning, and the device gracefully yields to pedestrians rather than forcing a pass.

To apply this in a last-mile fleet, operators should run controlled pilots, collect every incident data point, and adjust thresholds; late-day traffic scenes require tighter margins, slower speeds, and more conservative planning to maintain their mission while delivering more reliability.

Elad from the product team and Luke from ElevenLabs helped tune the perception and decision modules; their input, along with partners across scooter networks and other applications, improved interaction with traffic and pedestrians and refined how a driverless device handles curb ramps and sidewalk edges.

Practical deployment notes: start with small capacity routes, then scale to broader waitlists as capacity grows; deploying in stages helps track last-mile performance, reduce rash decisions, and iterate the product before expanding to the fleet and more applications, ensuring the product stays safe, friendly, and ready for real-world, driverless operation.

Delivery timing and routing: estimating last‑mile windows on urban sidewalks

Recommendation: set fixed last‑mile windows of 3–5 minutes per drop on urban sidewalks, and batch up to four parcels when traffic and pedestrian density allow. Use real‑time data to tighten or relax the window by 1–2 minutes as conditions shift, ensuring smooth handoffs and predictable service levels.

Context from CES 2019 shows Segway‑Ninebot’s Loomo debuting as a driverless device for sidewalk deliveries. The event, unveiled with movie‑style demos and broad media coverage, highlighted a hoverboard‑flavored platform aimed at urban par cels. Elad and PHIA teams followed the rollout, and the waitlist grew quickly as hundreds of operators evaluated the hardware’s practicality in real city traffic. Noting this momentum helps frame reliable last‑mile timing as more than a sci‑fi novelty; it’s a real, scalable ride‑hailing style problem for robotics and applications beyond a single device.

• Speed and maneuverability on sidewalks: plan for 0.8–1.6 m/s with a practical average near 1.0 m/s in moderate traffic, plus occasional slowdowns at crossings.
• Crosswalk and curb delays: expect 5–25 seconds per crossing, rising to 30–60 seconds in crowded blocks or during peak pedestrian flow.
• Parcel handling at curb: allocate 15–45 seconds per drop; four parcels add 1–3 minutes of dwell time per route.
• Traffic and pedestrian density: adjust windows by hour (morning rush, lunch lull, evening strolls) and by street type (arteries vs. side streets).
• Weather and surface conditions: rain or slick surfaces reduce speed by up to 15–25% and increase caution near curb edges.
• Energy and fuel considerations: electric drive keeps fuel costs low, but energy use rises with longer routes and frequent stops; factor battery range into routing decisions.

1. Define a routing objective that prioritizes reliability: minimize total route time while maintaining a safety margin of 15–30 seconds per stop.
2. Assign each sidewalk segment a speed profile and expected delay distribution, then compute arrival windows for all stops along the route.
3. Batch deliveries when window slack allows: combine up to four parcels on a single pass to reduce sidewalk presence and driverless idle time.
4. Incorporate dynamic rerouting every 60–120 seconds using live data on traffic density, pedestrian flow, and crosswalk wait times.
5. Embed safety constraints: keep 1.0–1.5 meters clearance from pedestrians and avoid high‑density zones during peak hours.
6. Validate against real deployments and adjust windows: use “noting” field data from hundreds of trials to tighten the model and reduce waitlist backlogs.

Example scenario: a 1.4 km urban corridor with four drop points, three curb stops and one mid‑block parcel handoff. Road speed on sidewalks: 0.9–1.2 m/s; crosswalk waits: 10–25 seconds; curb unloads: 20–40 seconds. Total estimated route time: 12–18 minutes, with a 2–3 minute buffer for variability. In a setup where most deliveries run in the late afternoon, planners can shift to a four‑drop batch with a 15‑minute window, lowering operational risk while keeping the waitlist in check.

Operational notes for every fleet: track actual drop times and adjust the model weekly. Media coverage of the debuting Loomo showed the potential scale of hundreds of deployments; use those learnings to calibrate service levels and share practical results with customers via clear parcel notifications and precise delivery windows. The overarching lesson is simple: timing must be as dynamic as the sidewalks are busy, yet predictable enough for operators and customers to rely on every day.

Power performance: battery life, charging cadence, and uptime expectations

Recommendation: Target a minimum of 8 hours of active operation per full charge on standard city routes, and implement a 4-hour charging cadence to sustain a fleet-wide uptime above 85% during peak shifts. Use field swaps to keep parcels moving and to avoid idle time in the line, which helps them deliver every hour.

Battery life hinges on payload, speed, and temperature. In urban conditions, expect 8–12 km per charge with a moderate load, translating to roughly 4–6 deliveries per charge. For a 20-robot fleet, allocate at least 4 charging bays that can handle 2–3 simultaneous cycles, so every robot returns to service within 90 minutes of a recharge window. This yields a realistic daily throughput for parcels and packages alike.

Charging cadence relies on smart chargers and battery health monitoring. Target 80% charges in 30–45 minutes when possible, with a full recharge of 1.5–2 hours if you have longer dwell times. Schedule recharges to align with routes and dwell periods so uptime stays high across the day. Also incorporate battery swap capability to reduce downtime during high-volume shifts, which helps the fleet scale.

The deployment is surrounded by sensors that monitor SOC, temperature, and the condition of the wheels. The company unveiled their fleet with a novel line of technologies, including modular batteries and wheel-driven propulsion, to minimize downtime. The hoverboard heritage informs efficient idle states, while the information dashboards track energy per parcel and per package, enabling precise comparisons to targets. In the market, drones compete for last-mile work, yet this bird-ground approach grounds reliability in the space where civilization operates, turning a novelty into a steady delivery routine and ensuring every route remains predictable.

Coco partnership impact: pilots, rollout plans, and collaboration models

Launch four pilots in four markets within the next quarter, leveraging segway-ninebot hardware and the Coco partnership to validate end-to-end delivery, including space for curbside and indoor staging, transporter routing, and package handling in dense urban space. Segway-ninebot unveiled its Loomo platform at CES 2019, illustrating direct applications for service logistics, says Coco executives. The objective is to measure capacity, service levels, and cost-per-delivery versus traditional courier models, then refine the rollout plan.

Rollout proceeds in staged waves: initial deployment in four corridors, then expansion to hundreds of routes across the sector over the coming months. The plan includes on-site service centers, spare parts within reach, and operator training. By sharing space across hubs, the Coco partnership reduces downtime and improves capacity utilization, then feeds data to calibrate delivery windows and pricing in real time.

Collaboration models center on four pillars: joint development with segway-ninebot and phia to translate concepts into field-ready features; licensing of transporter software; managed service partnerships with google to provide customer-facing service layers; and open APIs for local applications. Elad from phia leads the cross-functional team to turn concepts into scalable pilots, with hundreds of data points guiding decisions. The approach emphasizes cultural alignment with city rules and labor practices to ensure smooth adoption and rapid market fit.

The market for autonomous logistics continues to burgeon, with Coco targeting mid-market and enterprise sectors that demand reliable last-mile service. The collaboration aims to demonstrate applications that reduce package-handling time, improve traceability, and expand capacity without increasing street clutter. The proven benefits encourage this model to unlock space inside warehouses and on urban routes, while expanding the service portfolio and creating new revenue streams for all partners.