Services

What we do

We come to your floor, watch how the work is actually done, and then tell you three things: which steps are worth automating, which are not, and how to do it if you go ahead.

You can buy this on its own. Take the judgement and build it yourself, hand it to someone else, or hand it to us. We will not call something feasible in order to win the next contract.

What you get

A conversation, carrying what we concluded after seeing the floor.

We do not write a report. A well-written report is not the same as a correct judgement, and the hours spent writing it end up in the fee. What you need is to know whether to proceed and how — a conversation gets you there faster, and lets you push until you are satisfied.

You can keep asking. “Why isn’t this step worth doing?” “What happens if we do it the other way?” “What if the budget is half of that?” A report cannot answer follow-up questions.

Why it pays to talk first

The most expensive mistake in automation is not buying the wrong machine. It is automating a step that should have been left alone.

On any given line, usually only one or two steps genuinely justify the spend. The rest may be too low-volume, too variable, or fixable by changing how the work is organised — and putting a machine there makes the floor harder to adjust, not easier. Telling these apart takes a visit. It does not show up in a catalogue.

The reverse happens too. Some steps look impossible to automate until you approach them from a different angle, and then an off-the-shelf method already exists.

When this fits

• You know something on the floor is wrong, but not where to start
• A vendor has quoted you, and you want to know whether the proposal makes sense
• You want to size the investment without committing to a whole project
• Opinions inside the company differ, and you want a judgement with no stake in the outcome

What comes next

If you decide to proceed, we can take on the design and the build. That is a separate service — see Design and Build.

If you decide not to proceed, or to use someone else, that is a legitimate outcome of the conversation.

What we do

Turning an idea into a machine that runs crosses four disciplines: mechanical, electrical, software, firmware. We do all four ourselves.

On most projects the difficulty is not the depth of any one discipline. It is the boundaries between them. The mechanical designer does not know how stable the lighting has to be for the vision algorithm; the firmware engineer never sees the mechanical tolerances. Every revision then means another round-trip across three vendors. With all four in one team, those round-trips do not exist.

What the four cover

Mechanical design and fabrication. Fixtures, workstations, enclosures, mounts. From drawing through 3D printing and short-run machining.

Electrical and equipment integration. Wiring, PLC, sensor selection and hook-up, and communication between machines. Existing equipment that does not talk to anything, with data scattered across the floor, belongs here too.

Vision and software. Machine vision inspection, measurement and positioning, operator interfaces, data plumbing. Machine learning and large language model work is handled by the same people, not subcontracted.

Firmware and embedded. Flight control, motor control, sensor acquisition. The firmware on our own drones is handled by us.

When this fits

• Repetitive pick-and-place, assembly or inspection on the line
• Equipment already installed but not connected, with data scattered
• A machine or fixture you cannot buy off the shelf
• A project too small for the large system integrators to take on
• An idea about using AI on the floor, with no obvious first step

How it runs

1. Visit the floor and understand how the work is actually done
2. Propose a layout and equipment selection, with expected cycle time and investment
3. Procurement and build
4. Programming, vision calibration, on-site testing
5. Handover and operator training

If you are not yet sure whether to build anything, start with Consulting and Assessment. You can buy that on its own — we work out whether it is worth doing before anyone talks about building it.

No supplier lock-in

Robot brand, industrial camera, vision library, PLC and protocol are chosen per project. We hold no distribution agreements, so we have no inventory we need to move.

What we do

We design and assemble industrial drones for inspection and similar work. Our airframes have been flight-tested in an indoor studio, in factory aisles and on outdoor sites.

The reason for building our own is straightforward. Off-the-shelf airframes are designed for a general market, and against a specific operating environment or payload requirement they tend to fall just short. Building our own is how we close that gap.

When this fits

• Off-the-shelf airframes do not meet the constraints of the operating environment
• Flight is required indoors or in confined spaces
• A specific sensor or work payload has to be carried

Technical detail

Detailed specifications, the extent of in-house fabrication, and the current development stage are available on request.

What we do

We operate drones on site. Two kinds of work at present:

Solar arrays　Cleaning rooftop panels by drone. The aircraft carries the spray head and water is supplied from ground level through a tether, so nobody goes onto the roof. Dust and moss eat directly into output, and sending people up is not only slow but risks both falls and microcracks from footsteps.

Indoor facilities　Inspection of plants and enclosed spaces. Indoor flight has no GPS, which makes different demands of both the airframe and the pilot — one of the reasons we build our own.

Work is flown by personnel holding a Remote Pilot Professional Certificate from Taiwan’s Civil Aeronautics Administration.

When this fits

• A solar site needs periodic cleaning without sending people onto the roof
• A plant has high or hard-to-reach areas that need checking
• An indoor space needs inspection where GPS is unavailable

Technical detail

Site scale, throughput and pricing are quoted after assessing the conditions on site.