SharkNinja Co-op

Palindrome

In 2028, new EU eco-regulations will require vacuum cleaners to be twice as energy-efficient. At SharkNinja, the strategy was to design a vacuum that could clean just as effectively on the backward stroke as on the forward stroke. That's how the Palindrome vacuum concept was created.

Palindrome vacuum nozzle prototype

The main element determining reverse-stroke performance is the vacuum nozzle, and during my time at SharkNinja I worked on improving existing designs and developing new concepts for future products. My personal goal for Palindrome was to develop a working proof of concept and establish a design direction that the R&D team could build on and iterate after my co-op ended.

Goals & Limitations

Project Goals

  1. Get maximum debris cleaning. Current intake designs get ~70% pickup on the backward stroke vs. ~100% on the forward stroke — is it possible to get 100% in both directions?
  2. Create a maximum efficiency nozzle. Current tests take 3–5 strokes — can this nozzle design achieve parity scores in 2 strokes?

Limitations

Conventional nozzles have a brushroll before the suction inlet that agitates the material and lifts debris off the surface, which is then sucked up. Behind the suction inlet is a bristle strip that contacts the ground to create a seal and maximize suction.

Current designs have three problems:

  1. In the reverse stroke, the rear seal is opened, creating suction loss and less effective reverse cleaning
  2. The rear seal does not open enough, causing a buildup of debris ("snow piling")
  3. The rear seal opens too much, causing debris to hit the brushroll and shoot back toward the user ("spit-back")
Forward stroke
Backward stroke — fails to pick up debris due to snow piling
Directional data using current production vacuum
Debris Type (20g) Debris Size Debris Forward Stroke Backward Stroke
Sand Fine Dense 98.75% 47.9%
Coffee Fine Light 99.15% 99.05%
Rice Large Dense 99.5% 91.5%
Crushed Cheerios (7g) Large Light 95.57% 1.24%

Current design struggles with fine + dense and large + light debris types due to snow-piling, especially on the backward stroke.

Reverse Intake Vent

My initial concept was based on an existing nozzle. I focused on eliminating spit-back and separating the air pathways for the forward and reverse strokes.

Section view CAD diagram of the nozzle base, labeling the suction inlet, the airpath, and where the bristle strip is inserted

I designed a nozzle base that splits the airflow between forward and reverse directions. The reverse airpath bypasses the bristle strip to avoid compromising the seal and eliminating any spit-back.

Initial tests with split airpath:

  • Split airpath causes significantly weaker suction in both forward and backward directions
  • Lack of agitation in the backward stroke led to no pickup

Tests with 100% airflow to the back vents:

  • Back intake still had limited success
  • Could not eliminate snow buildup at the nozzle ends despite multiple design iterations

Learnings

Palindrome: Mirrored Nozzle

From the reverse intake vent, I realized that effective cleaning would require agitation in both directions. This led me to design a mirrored nozzle featuring a brushroll, intake, and bristle strip on each side. The main challenge was designing an alternating valve that could redirect the airpath based on the direction of the stroke.

Diagram comparing the mirrored nozzle's forward and rearward airpath configurations, labeling the brush roll, nozzle housing, static squeegee, and alternating door mechanism

Palindrome (w/o brushroll) vs. Current Nozzle (PowerDetect)

Palindrome
PowerDetect

The prototype Palindrome performs identically in either direction, while the PowerDetect struggles and "snow-piles" on the reverse stroke.

Alternating Valve Development

V1
Diagram comparing the alternating valve's forward and rearward airpath configurations
  • Debris ingress big concern at pivot axis
V2
  • Redesigned slot geometry to minimize dust ingress
Diagram showing airpath direction through V2's redesigned slot geometry in the forward and rearward configurations CAD screenshot of V2's redesigned valve disc slot geometry
  • Mechanism would not allow for rubber seal: no radial force that could create a seal
V3
  • Added additional CAM slot to create a lifting motion for the door
CAD screenshot of V3's valve disc slot geometry
  • Adds motion to the door (pops up then across)
  • More force needed to articulate the system
V4
  • Changed to a straight CAM to reduce friction
V4 slot geometry with the added CAM slot circled in red, which creates a lifting motion for the door
  • Debris ingress is noticeable after some testing with sand
  • Smoother/easier articulation
  • Full articulation; open or closed
V5
  • Dual nozzle attachment using two nozzles back-to-back and connected to FW 3.0 (Light vacuum)
  • Inconsistent articulation: not enough friction between ground and pivot (using hot glue as grip)
  • Not enough suction to pick up debris without a brush roll
  • Partial airflow leakage through closed valve
V6
  • Reduced number of components
  • Internalized moving sections to reduce air leakage and improve prototype robustness
  • Reduced overall size
  • Adapted the nozzle to a PowerDetect for higher suction
CAD render of the fully assembled V6 valve, adapted for a single PowerDetect nozzle Front CAD view of the V6 valve integrated into the PowerDetect nozzle housing Detail view of V6's inward-facing slot design, circled
  • Bottom-facing slot will accumulate debris
  • Increased push-pull force from feet with no change in the y-axis

Next Steps for the R&D Team