Research

Research

Norcium is a research shop as much as a product shop. These are the areas we're actively exploring — some already shipped in Peek, some still in the lab.

Active pillars

Long-running research threads that inform everything we build.

Shipped in Peek

Ambient & Voice Computing

Wake-word activation, proactive unprompted commentary, and calendar-aware assistants that act without being asked — interfaces that fade into the background until they're useful.

Shipped in Peek

Generative Audio

Real-time synthesized soundscapes generated from live data (not pre-recorded files) — audio that reacts continuously to changing conditions.

Shipped in Peek

Augmented Reality

Compass-anchored AR overlays fusing sensor data (wind direction, sky position) with the live camera feed.

In the lab

Nowcasting & Prediction Modeling

Minute-by-minute precipitation nowcasting and personal correlation modeling (symptoms/mood vs. weather) — small models, tightly scoped, run close to the user.

In the lab

Behavioral Gamification

Streaks, quests, and prediction games studied for genuine engagement lift versus novelty decay — what keeps people coming back after week three.

In the lab

Accessible-by-Default Design

Motion, voice, and glass-morphic UI systems built with VoiceOver and Reduce Motion as first-class constraints, not retrofits.

New research

Desalination & Water Tech

Turning seawater into drinking water — reverse osmosis, energy recovery, and where the economics of fresh water are heading next.

Read the full research note →

Research note: Desalination

How ocean water becomes drinking water, what it costs today, and where the technology is heading.

How it works

Reverse Osmosis (RO) is the dominant method (~56% of the market). Seawater is forced under high pressure through a semi-permeable membrane that blocks salt ions, leaving fresh water on one side and concentrated brine on the other — more energy-efficient than boiling water.

Thermal methods (still common in the Gulf): Multi-Stage Flash (MSF) heats seawater and flashes it into steam across successive lower-pressure chambers; Multi-Effect Distillation (MED) reuses the latent heat from each stage to boil the next, making it more efficient than MSF.

Emerging methods — electrodialysis, forward osmosis, and membrane distillation — are still mostly in development and haven't displaced RO at scale.

Energy recovery devices (ERDs) are the key efficiency trick: they capture pressure energy from the leftover brine stream and feed it back into incoming seawater, recovering up to ~70% of energy that would otherwise be wasted — cutting RO energy use by up to 60% since the 1990s.

Industry & data (2026)

Global market: ~$20.8B in 2026, projected to reach $38–59B by 2032–2034 (CAGR ~9%).

Largest plant: Ras Al Khair, Saudi Arabia — ~3 million m³/day, built for ~$7.2B by SWCC.

Operating cost: modern large RO plants produce water for $0.40–$0.80 per m³; Saudi Arabia's newest projects hit ~$0.50/m³ helped by cheap solar power; Dubai's Hassyan plant targeted a record ~$0.31/m³.

Capital cost: roughly $1,000–$2,500 per m³/day of capacity — megaplants (>500,000 m³/day) run 25–40% cheaper per unit than small plants.

Biggest operating expense: powering the high-pressure pumps — 35–45% of OPEX.

Heaviest adopters: Saudi Arabia, UAE, Israel, Spain, Australia — increasingly California and India.

Where it's heading

Membranes: salt rejection now above 99.7%; graphene-oxide membranes in development promise better rejection at lower energy cost.

Solar desalination breakthroughs: new passive solar designs avoid producing toxic brine waste — a major 2026 research theme.

AI-driven plant optimization: real-time membrane fouling detection and energy tuning.

Zero-liquid-discharge (ZLD): turning waste brine into usable salt/mineral products instead of dumping it back into the ocean — addresses the biggest environmental criticism of desalination.

Sources: Blackridge Research, Grand View Research, Elemental Water Makers, ScienceDaily

Stay in the loop

Research notes, occasionally

No spam, no schedule — just the occasional note when something in the lab is worth sharing.