LOBO and Fleet

LOBO is a solar security tower with its own power, cellular connection, cameras and deterrence. It arrives on a trailer, rolls into place and is watching within a day. Fleet is the app that runs it.

What's on the tower

LOBO from the front with the mast raised, with numbered markers on the equipment box, cameras, mast, solar panel and rolling base.
LOBO with the mast raised. Hover or tap a number to find the part.
  1. Equipment box

    Holds the cameras, the talk-down horn, the light and the electronics that run them.

  2. Panoramic cameras

    360° coverage around the tower.

  3. PTZ camera

    Long-range optical zoom, so an agent can look closer.

  4. Crank mast

    Raises the cameras up to 20 ft. It comes down to move.

  5. Solar panel

    Powers the tower. No site power or wiring.

  6. Rolling base

    Batteries and charge controller on pneumatic tires.

How it responds on site

LOBO reacts where it stands: cameras turn to follow, the lights flash, a loud siren sounds and an agent speaks through the horn speaker. Anyone there can tell it's watching.

  • Follow

    Cameras that track

    On current builds, tracking cameras turn to follow people and vehicles, so anyone on site sees the tower watching them.

  • Sound

    A loud siren

    A siren and an alarm message through the tower's horn speaker, loud enough to hear across the site.

  • Light

    Police-style lights

    Lights and a strobe on the mast, on demand or on a schedule.

  • Speak

    Live talk-down

    An agent, or you from your phone, speaks to the person on site through the horn speaker.

E911 from the tower

Each tower has its own phone line, registered for E911 at its site's street address. A 911 call placed from the tower reaches local dispatch with that address.

Calls to your contacts

Ring your site contacts from the same control panel.

The Fleet app

The app our customers use to watch their sites, on phone and desktop, and the software we run our own towers on. Every tower appears on a 3D map where it stands, with its live status.

Fleet's 3D map at a logistics yard: a large warehouse in photorealistic 3D, with each tower labeled L096 to L099 where it stands and a beam rising from it.
Fleet at a deployed yard in South Texas, each tower labeled where it stands. Map imagery ©2026 Google.

Multi-layer AI

Each event passes through layers. Detectors on the tower find people, vehicles, drones and gunshots. A larger AI model then judges the scene the way a guard would, knowing the site's hours, its guard and its streets, and decides whether a person needs to see it.

  1. Layer 1

    Detection on the tower

    Live

    Camera AI and our on-board model fire on people and vehicles. The tower's microphone listens for drones and gunshots.

  2. Layer 2

    Scene review with site context

    Rolling out

    A state-of-the-art vision-language model looks at the event's pictures in order, with the site's info sheet (hours, who belongs there, what to watch for) and an aerial map of the site and its streets. It writes what it sees and why, then clears the event or forwards it. Scheduled looks at every camera are next.

  3. Layer 3

    Human agents

    Live

    Agents act on what needs a person: siren, lights, talk-down, and calls to your contacts or the police.

Layer 1: detection

  • Night view from a tower camera at a warehouse lot: a person walking with a flashlight, boxed by the detector and labeled Person, 90 percent.

    People

    Live

    On every camera, day and night.

  • Daytime view from a tower camera: an SUV pulling into a warehouse lot, boxed by the detector and labeled Vehicle, 96 percent.

    Vehicles

    Live

    Cars, trucks and equipment entering or moving on site.

  • Illustration: an audio spectrogram from 0 to 8 kHz with a drone's stack of harmonic lines, boxed and labeled Drone.

    Drones

    Beta

    The tower's microphone listens for small drones.

  • Illustration: an audio spectrogram with two gunshots, short bright spikes across every frequency, each boxed and labeled Gunshot.

    Gunshots

    Beta

    The same audio model listens for gunshots.

People and vehicles: real detections from a tower at a logistics yard in South Texas, with the face and plate blurred. Drones and gunshots: illustrations.

Layer 2: scene review

Real reviews from our sites, in the model's own words. Each pair starts from the same detection, a person at night, then a vehicle in the dark, and ends with a different answer.

  • Night view of a lit equipment yard: a person in an orange vest at a guard shack beside a portable toilet, a white SUV parked nearby and a small dog.
    Clear11:47 PM · guard hours

    The night guard at the post

    “Person in orange hi-vis vest, no hard hat, standing at the guard shack steps beside the porta-potty, white SUV parked by the shack, small dog nearby, at 23:47–23:49.”

    Why“The person matches the overnight guard described on the info sheet: guard hours, orange hi-vis with no hard hat, staying at the shack and porta-potty, with the white SUV parked by the shack. Nothing suggests intrusion or tampering.”

  • Night view down a fenced path on a vacant property: one person in a dark jacket and light pants, carrying a bag, walks toward the camera.
    Forward11:17 PM · vacant site

    Someone who isn't supposed to be there

    “Person in dark jacket and light pants, carrying a bag, walking on the paved path between fences toward the tower at 23:17 on a vacant site.”

    Why“The info sheet says no one should ever be on this vacant site, and a person is on the property at night between the fence lines. Trespassers are on the watch list, so an operator should talk them down.”

  • Pre-dawn view of a warehouse entrance: a pickup with its taillights on drives past the glass front doors, with a red pickup parked at left.
    Clear5:54 AM · before opening

    Staff arriving before opening

    “At 05:54 a pickup drives past the main entrance, the guard's SUV sits by the booth, and a lot full of staff cars is filling up.”

    Why“The info sheet says staff start arriving about 5:15 AM. Six minutes before opening, a full employee lot and a single pickup moving on the drive by the entrance, with the guard's SUV at the booth, fit normal staff arrival.”

  • Night view from a tower camera: a red sedan on a closed site's drive by the building entrance, brake lights on, in a haze of tire smoke over circular skid marks.
    Forward2:03 AM · site closed

    A car doing doughnuts at 2 AM

    “Red sedan on the property's paved drive by the building entrance at 02:03 with brake lights on, tire smoke around it and circular skid marks on the concrete.”

    Why“An unknown vehicle is on the property in the middle of the night while the site is closed. The tire smoke and circular marks suggest drifting or doughnuts, and the info sheet asks us to report vehicles on the property after hours.”

Real events from October 2026, quoted from the model's summary and reasoning, trimmed and with tower names left out; faces, plates and signs blurred. Layer 2 runs alongside human agents while we measure it.

Scheduled looks

In development

Detectors fire on people, vehicles and animals, never on smoke or water. So Layer 2 will also look at every camera on a schedule, event or not, and flag smoke, fire, water where it shouldn't be, an open gate, or anything else out of place.

  • Daytime view across a warehouse lot: a thick black smoke plume rises behind the building into a clear sky.
    Forward1:12 PM · working hours

    Smoke behind a warehouse

    “Thick black smoke plume rising in the upper left of the frame, behind/beyond the long warehouse building …”

    Why“Clear sunny sky, so this is not fog or a lens artifact. Source is not visible; it may be behind the building or off-property. Needs immediate verification.”

  • Sunny afternoon in front of an office building: a wide sheet of water covers the drive around a planted island with yellow bollards, while the pavement beyond is dry.
    Forward3:53 PM · Sunday

    Standing water on a sunny day

    “Large sheet of standing water covers much of the parking lot and drive in front of the building, reaching the curbed planter by the green utility box and yellow bollards.”

    Why“The sky is sunny and the surrounding pavement is dry, so this is not rain runoff. The water around the base of the planter near the utility box looks disturbed, as if it is flowing out there. This may be a broken water line or irrigation leak.”

  • Overcast view across a parking lot, with grey, blurry patches where water drops sit on the camera lens.
    Clear1:52 PM · after rain

    Water on the lens, not smoke

    “Water drops or smudges on the lens make grey, blurry patches in the center and right of the frame. They partly obscure the view, but most of the scene is still visible.”

    Why“These are on the lens, not smoke. Overcast sky and wet pavement suggest recent rain. No standing water or flooding is visible.”

Replayed on recorded footage and quoted from the model's findings. Smoke: a warehouse fire in Dallas on October 22; no camera fired on it, and a human agent saw it and called 911. Water: an overflowing drain at a business park in Georgetown, TX on January 4, which a human agent reported. Wet lens: a tower after rain on October 24.

Live view and controls

Open a tower to watch its cameras live. Talk-down, siren, lights and calls are on the same screen, with the tower's events beside them.

Illustration of Fleet. The camera frames are from a deployed tower; the unit list and events are examples.

Software updates

We write the software on the tower and in Fleet, and send updates to every tower over its cellular link. Here is what is live and what is next.

Illustration of on-tower acoustic detection: a drone's harmonic lines and a gunshot's broadband spike.
  1. Live

    Live view and controls

    Every camera, plus talk-down, horn, lights and calls, on phone or desktop.

  2. Live

    3D and 2D maps

    Every tower on the map with its live status.

  3. Live

    Recordings timeline

    Scrub recorded video from the tower's own storage, with detections marked.

  4. Live

    Your team and your systems

    Read-only viewers and Microsoft or Google sign-in. Streams into your own VMS and an API, on request.

  5. Rolling out

    AI review

    A second, multi-frame look at each event that judges people the way a guard would, before it reaches a human agent.

  6. Beta

    PTZ follow

    The PTZ turns to a person the panoramic cameras find and keeps them in frame.

  7. Beta

    Gunshot detection

    An audio model on the tower listens for gunshots and logs them beside the camera events.

  8. Beta

    Drone detection

    The same on-tower audio model listens for small drones.

Specifications

Standard configuration

Coverage
360° from multiple cameras
Zoom
PTZ camera with long-range optical zoom
Night
Infrared or low-light cameras
Resolution
Up to 4K, varies by camera
Recording
Up to 2 TB on the unit
Height
Up to 20 ft
Power
Solar, with up to 4 days on battery alone
Connection
Built-in 5G/LTE cellular
Deterrence
Siren, talk-down, police-style lights
Microphone
External mic on current builds, for live audio and acoustic detection
Access
Live view on phone or desktop

Specifications describe our standard configuration and vary by unit and site. Performance depends on sun, weather and cellular coverage at the site. Battery figure: current build at the fleet's median draw, October 2026. Audio is recorded only where the law allows.

Get a quote or a demo.

Tell us about your site. We'll get back to you within 48 hours.

Glenn Wolfe
Co-founder
[email protected]
David Gonzales
Co-founder
[email protected]
I'm interested in