The Tech Blog

I recently read about lithium-air batteries and got excited by their potential. Their theoretical energy density approaches petrol. Let’s discuss why.

Today’s commercial lithium-ion cells typically offer 160-300 Wh/kg. Specialized silicon-anode cells reach 450 Wh/kg. Petrol packs around 12,000 Wh/kg of chemical energy. Lithium-air offers 11,400 Wh/kg on a lithium-only theoretical basis, or about 3,500 Wh/kg including the oxygen consumed. Neither figure describes a finished rechargeable battery.

Lets talk chemistry. A lithium-air cell uses oxygen from outside. Lithium metal releases electrons into your circuit, while Li⁺ ions cross the electrolyte. At a porous cathode, oxygen, Li⁺ and electrons form lithium peroxide (Li₂O₂). Charging must reverse this reaction and release oxygen. It effectively breathes in like us.

Earlier this year, researchers in Japan demonstrated 740 Wh/kg from a 5.2 g experimental stack using dry air. Another study reported almost 799 Wh/kg in a prototype pouch cell. Separate limited-capacity cycling tests in the first study achieved around 68% round-trip energy efficiency and fewer than 100 cycles.

There are real hard problems in the tech though. Lithium peroxide clogs cathode pores and blocks oxygen. Moisture and CO₂ in normal air cause side reactions. Reactive oxygen damages the electrolyte, while lithium metal must endure repeated charging. Even lithium-ion cells age, need thermal control and risk lithium plating during rapid charging.

I think drones and aircraft may benefit first, because every kilogram matters. Cars will need more power, longer battery life and lower costs. Lithium-air is still at a very early stage, and there is a long way to go. But I believe we could one day build batteries that deliver as much usable energy per kilogram as petrol. That would be a huge step towards leaving petrol behind and moving to a world powered by renewable energy.

I was recently going through IEEE 802.11bf, the Wi-Fi sensing standard published last year. It kind of formalised procedures for Wi-Fi sensing. I think it’s worth diving into how it works.

So a Wi-Fi signal reaches the receiver through several paths. Some travel directly, while others reflect off walls, furniture and people. These copies combine at the receiver. When you move, the path lengths change, shifting their phase and amplitude. Wavelength is about 6 cm@5GHz. In an ideal out-and-back reflection, moving a person 1 cm towards the radio changes the path by 2 cm, producing a 120° phase shift. That measurably changes the received signal.

Wi-Fi receivers already estimate Channel State Information (CSI) to decode data. Wi-Fi splits a channel into many subcarriers. CSI estimates how each one’s amplitude and phase have changed during transmission. That gives much more detail than a signal-strength (RSSI) reading. By tracking CSI over time, software can detect motion and, under controlled conditions, infer gestures or even breathing. Unlike cameras, radio sensing may operate through walls, raising privacy concerns.

Now, can your home router reconstruct a 3D map of your room? In principle, you can estimate locations, but a typical router cannot produce a detailed 3D scan. You need multiple radio links or antennas, bandwidth and calibration. At 60 GHz, wider channels help resolve finer details, but blockage becomes an issue. Your router being Wi-Fi 6 or 7 alone doesn’t enable sensing. The chipset and firmware must expose useful channel measurements, with software to interpret them.

The company Plume already offers Wi-Fi motion alerts, and researchers have demonstrated 3D human pose tracking with multiple receivers. But movement from fans or pets can trigger false alarms. A person sitting still is much harder to detect than one walking. CSI reveals that something changed. Deciding what caused it remains the hard part. As sensing improves, maybe one day Wi-Fi could help automate lighting and alert caregivers to possible falls of elders without cameras. Future is great for this tech.