Two products, one pipeline. A picket of passive counter-UAS sensor nodes you place forward, and a compliance checkpoint that never touches identity. Both run on commodity hardware, publish open messages, and keep a record you can audit.
Small drones are now the thing that kills soldiers at the front, and the layered air defense has a hole where they fly. Radar sees what is big. RF detection sees what talks. A Group 1 quadcopter on a fiber line, or on an autonomous terminal run at 150 meters, is neither. It is small, it is low, and by the time someone sees it there is nothing left to do but duck.
Tripwire is built for low, slow, small threats that fly below the radar horizon: Group 1 quadcopters and FPVs, including fiber-guided, and Group 3 one-way attack UAS of the Geran class. The bigger it is, the farther out a node picks it up. Today the detection layer at that range is a soldier’s eyes and ears. Every interceptor in the Replicator-2 class is only as good as the track it is launched on, and close in, where it needs that track most, there is no track. That is the gap we built for. It sits under Formation-Based Layered Protection, and it sits under the systems you already own. It does not compete with them.
Tripwire is a passive sensor node that puts a 3D track on a small drone. You place nodes forward, where you expect them to come from, and the first node to detect one tells everyone behind it. Each node waits in a passive, low-power state and looks only when its detector picks something up and gives it a sector and a rough range. On a hit, the node wakes a near-infrared camera and an active optical ranging method of our own design, patent pending, and puts a real 3D position on the target, not a bearing, in about a second. The track is published as Cursor-on-Target, so anything that reads TAK can act on it, and it can go straight to whatever is going to shoot the drone down.
Cues run both ways. Radar, an RF detector, an IFF non-reply, or higher echelon C2 can point a node into a sector and get back a confirmed or rejected track. Nodes share a GPS or PTP clock, so two nodes that see the same target fuse into one tighter fix. The node radiates nothing until it detects a target, then emits near-infrared light for a fraction of a second per frame. It is EMCON-safe by design, and it is cheap enough to lose. A platoon can field a dozen and not mourn one.
Reach is a matter of where you put them. Each node covers a bubble a few hundred meters across. A dozen nodes along a tree line or an approach corridor is a kilometer of frontage, and warning time is distance divided by drone speed. A quadcopter at 15 to 20 meters a second covers a kilometer in about a minute, so nodes placed 2 kilometers out buy about two minutes and nodes 10 kilometers out buy eight to ten. A track is a Cursor-on-Target message of a few hundred bytes, so it rides on a mesh radio, a LoRa link, a cable, or a tactical radio you already issue. A node transmits nothing until it has something to say.
It is low-collateral by design. The same optics that find a drone also see a person, and an oversized return shuts the emitter off within one frame. The node carries a co-aligned visible marker to designate the target for a gunner, so the first effect on target can be a bullet or a net instead of an RF attack that also jams your own radios.
Nothing exotic in the parts list. Repairable at unit level, updated over the air.
We are a sensor and a cueing layer. We do not make the C2 and we do not make the effector. We feed both, in the message formats they already speak.
We are building to the Modular Open Systems Approach from the first unit. Every interface on the node is open and documented: Cursor-on-Target out, cues in over the same channel, time from GPS or PTP, power and data over standard connectors, and a node record in open formats. Our roadmap aligns the node to the SOSA Technical Standard’s sensor-module conventions and to the joint C-UAS interface standards as they are published, so a program office can swap us in, or swap us out, without touching the rest of the stack. We would rather be one module in your architecture than a system you have to build around.
Every base, depot and flight line already knows who is walking through the door. What it cannot show is whether that person was in the right gear, on that shift, at that gate. WorkCheck answers that question at the door, in about a second, and writes the answer to a record that holds up, without touching anyone’s identity.
It uses no face recognition and stores no image. No biometric template ever exists, so there is nothing to protect and nothing to subpoena. The check runs on the device and the record is a hash, so it works with no network at all. Running air-gapped is the default, not a special mode. The console signs in through the base identity provider, so CAC and PIV work without a new password, and for regulated tenants publishing a rule takes two people, one to draft and one to approve.
We are a U.S. small business in Utah, 40 minutes from Hill AFB. Both products are built and tested here. SAM registration is in process, and we are eligible for SBIR and STTR, xTech, OTAs through the consortia, and CRADAs. We have applications in process with the Army and other DoD components. We work with university research partners for optical characterization and edge inference.
If you have a requirement, send it. We answer RFIs within a week with a one-page capability summary, a quad chart, and a bench video. If you have a range and a target, we would like to come out.
Every verdict and every track is produced on the device. No cloud in the loop. Patent pending on the Tripwire sensing method.