Cleaning the glass on a forty-story office tower in a major US city is a labor-intensive, high-risk operation. A suspended scaffold crew rigs from the roof, works down the facade at a few floors per day, and repeats four to six times a year on a Class A building. Falls from elevation remain among the leading causes of death in construction and building maintenance, and the insurance premiums attached to that risk are a permanent line on every building services company's income statement.

A drone that sprays and scrubs the same glass from the air removes the person from the exposed position. That is the entire commercial argument, and it is a strong one. The engineering required to deliver it is harder than the pitch deck suggests.

Where the Money Actually Is

Facade cleaning on a high-rise typically runs one to three dollars per square foot of glass depending on the market, access difficulty, and frequency. A large tower can carry a six-figure annual cleaning contract. The costs a drone attacks are not primarily labor hours; they are the fixed costs of working at height: rigging and de-rigging suspended equipment, permits and street closures, fall protection compliance under OSHA 29 CFR 1926 Subpart M, the specialized insurance, and the schedule constraint that scaffold work stops in wind that a drone can still handle marginally better.

Building owners also care about disruption. A scaffold outside a trading floor window for two weeks is a complaint generator. An aerial system that covers the same elevation in two days sells on that alone, which is the same argument that moved facade surveys onto aircraft in drone infrastructure inspection.

Tethered by Necessity

Almost every serious window-cleaning drone is tethered, and for two independent reasons.

The first is water. Cleaning consumes water continuously, and carrying it is hopeless: a gallon weighs about 8.3 pounds, and a multirotor able to lift a meaningful tank spends its entire payload budget on liquid it will exhaust in minutes. A hose from a ground or rooftop unit gives unlimited supply at the cost of hose weight, drag, and a reel system that must pay out and retract under control.

The second is power. A battery-only aircraft flies for twenty to thirty-five minutes and then stops for a pack swap; a facade job wants hours of continuous work. Running power up the same umbilical solves that, using the high-voltage transmission and step-down architecture explained in tethered drone systems. Sending several hundred volts up thin conductors and converting at the aircraft keeps copper weight manageable.

The tether is then the dominant design constraint. Its weight and drag change continuously with deployed length, its catenary behaves differently in wind, and it can snag on balconies, sunshades, window washing anchors, and antenna masts. Tether management systems that maintain constant tension rather than fixed payout are what separate a demonstration from a product.

The Cleaning Head

Two approaches dominate. Pure water systems use deionized or reverse-osmosis water sprayed at pressure and allowed to dry without squeegeeing, since water with no dissolved solids leaves no spotting. This is already the standard in ground-based water-fed pole cleaning and it adapts well to aircraft because it removes the need for physical contact and drying passes.

Contact systems add a rotating brush or a soft pad pressed against the glass. They clean more aggressively, which matters for bird droppings and mineral staining, but they introduce a control problem: the moment the aircraft touches the building it has an external force and reaction to manage. Solutions include compliant contact heads, a passive standoff frame, and control modes that deliberately use the facade as a reference. Nozzle and flow design carries over directly from the coverage work in agricultural spray drone development, where uniform deposition at a controlled standoff is the same problem in a different orientation.

Flying Close to a Building

The aerodynamic environment beside a tall facade is hostile. Wind accelerates around corners and forms vortices, vertical channeling can produce strong updrafts, and the aircraft's own downwash reflects off the glass and destabilizes it at close standoff. Practical designs hold a standoff distance and spray across it rather than hovering nose-in at inches.

GNSS is unreliable beside a building that blocks half the sky and reflects the rest. Position relative to the facade has to come from onboard sensing: lidar or ultrasonic ranging to the glass, visual tracking of mullions and window frames, and inertial fusion to ride through dropouts. The sensing architecture is the same family described in drone obstacle avoidance and GPS-denied navigation, tuned for a flat, specular surface that confuses many optical sensors.

The Regulatory Reality

Commercial operation runs under FAA Part 107 with a remote pilot certificate. Three provisions bite in this application.

  • Operations over people. A city sidewalk beneath a tower is exactly the situation the rule addresses. Meeting one of the operational categories, or closing and controlling the area below, is a precondition. Mitigation hardware such as the systems in parachute recovery and flight over people forms part of that argument.
  • Altitude. The rule allows flight within 400 feet of a structure's uppermost limit, which accommodates most towers, but airspace authorization is still required in controlled airspace, and downtown cores are usually controlled.
  • Visual line of sight. The pilot must see the aircraft. On a tall building this often means positioning an observer on a roof or an adjacent structure.

Local rules add another layer. Many cities require permits for sidewalk closures and for any suspended work over a public way, and building management will want proof of insurance naming them. Water runoff is a real issue too: wastewater running off a facade onto a street is regulated in some jurisdictions, and containment or scheduling around it belongs in the operating procedure.

Building the Business Case Honestly

A tethered facade-cleaning system is a capital product costing well into six figures, sold to building service contractors who will compare it against a scaffold crew they already own. The credible pitch is throughput per day plus reduced exposure, not labor elimination, because the system still needs a pilot, a ground technician, and a water plant. Systems selling well today target repeat institutional customers with large glass inventories, the same buyer profile that adopted the aircraft in construction site drones.

Develop the System With Us

Projects House develops tethered and payload-carrying aircraft for US clients through a global engineering and manufacturing network, from tether and power architecture through spray head design, facade-relative navigation, and field trials. Send your target building type and cleaning method through our contact form.