On‑demand energy, predictable duty cycles, and standards you can deploy today.
Most “battery‑free” ideas sound great—until you have to make them reliable. Reliability, in our world, means a sensor that delivers a clean measurement exactly when you ask for it, in the places you care about, for years, without maintenance rituals. That’s why at Kliskatek we prioritize far‑field RF (sub‑1 GHz UHF / RAIN RFID) over ambient sources like indoor light or vibration when reliability is the top requirement. Readers bring energy on demand, and the air interface is mature and global.
Thesis: If your success depends on getting a reading at a specific moment (not “whenever nature cooperates”), RF beats natural sources most of the time—especially indoors and in industrial settings.
What “reliable” means in practice
- On‑demand energy: A RAIN RFID reader transmits UHF power; the passive tag harvests that RF, wakes, measures, and backscatters the data—no battery required. The process is defined by EPC Gen2 / ISO/IEC 18000‑63 and works across regions.
- Predictable read windows: RAIN systems routinely inventory hundreds of tags per second over several meters; with tuned antennas, 10–20 m is achievable for ID tasks. That throughput and range create short, dependable power/comm windows for sensing.
- Standards & scale: The ecosystem (tags, readers, middleware) is large and interoperable; Gen2v3 keeps improving inventory/select and user/TID data access.
RF energy harvesting 101 (the 30‑second version)
- Reader → Tag: Reader radiates; tag’s rectenna turns RF into DC; the tag modulates the reflection (backscatter) to send data back.
- Why sub‑1 GHz helps: Lower frequency means lower free‑space path loss for a given antenna gain (per Friis), which improves power transfer headroom at a few meters.
- Regulatory reality: Enterprise readers commonly support up to ~33 dBm EIRP (region‑dependent), giving you practical energy budgets at 3–5 m.
Why “natural” sources disappoint when you need predictability
1) Indoor light (PV)
Indoor photovoltaics can be excellent for µW‑level devices near constant, known lighting, and efficiency keeps improving (perovskite/OPV/TMD cells). But reliability issues remain:
- Low, variable irradiance & spectra: Lux is not irradiance; lux‑based estimates mislead, and there is no universally accepted standard for indoor PV test conditions—making repeatable power forecasts hard.
- Power density limits in real rooms: Even careful studies show indoor PV comfortably powers µW‑class nodes, while tens of mW sustained is challenging under typical mixed lighting without careful placement.
- Research momentum ≠ field uniformity: 2023–2025 reviews highlight strong lab progress (and even dual use as VLC receivers), but deployment still hinges on environment‑specific lighting control.
Bottom line: Great when illumination is guaranteed and modeled; unreliable when lights go off, dim, or their spectrum shifts.
2) Vibration
- Resonance dependence: Most VEH designs peak at a narrow frequency; ambient vibrations are random/broadband and change with machine states, so harvestable power can swing by orders of magnitude.
- “Machine‑on” assumption: No motion = no energy. Even promising industrial demonstrations that power low‑duty sensor nodes show intermittent energy accrual (e.g., ~138 mJ/day on a rotary pump) and require meticulous matching.
Bottom line: Useful where vibration is steady and well‑characterized; unreliable as a universal baseline.
Why RF is different: you control the energy
- You bring the field. With a reader dwell, you decide when energy is present; the tag then executes a short, deterministic transaction.
- Throughput creates options. High read rates let you retry quickly, collect diagnostics, or inventory many sensors within a pass.
- Physics on your side. At sub‑1 GHz, path loss is friendlier; with reasonable gains and 33 dBm EIRP, the few‑meter region offers workable harvested power for a single measure‑and‑report cycle.
A practical (vendor‑agnostic) design recipe
- Design for a short, single‑pass cycle: Boot → energy check → measure once → serialize → backscatter → sleep. Keep it idempotent. (Intermittent power demands it.)
- Budget usable capacitor energy: Size storage for Vmax→Vmin (not just ½·C·V² at Vmax). Then set brownout and write thresholds with margin. (General RF/EH practice; aligns with Gen2 access timing.)
- Exploit diagnostics over the link: Include sequence counters and voltage at start/end so operations can distinguish “low‑energy window” from sensor faults on the backend. (Plays well with standard readers and parsers.)
- Tune antennas in‑situ: Industrial environments are multipath‑rich; enclosure + materials matter. Field reality beats lab air.
When “natural” still wins
There are real cases where indoor PV or vibration is the right choice:
- Controlled lighting environments (e‑paper displays, shelf labels) where illumination is guaranteed.
- Always‑on machinery with stable spectral content at a known frequency, allowing well‑tuned VEH and modest duty cycles.
The best systems sometimes go hybrid—we start from RF because it gives you a reliable baseline even when nature is uncooperative. (We’ll explore hybrids later in the series.)
Field‑ready checklist
- Reader dwell planned for the weakest coupling spot you expect.
- Storage C sized for Vmax→Vmin and a single deterministic cycle.
- Antenna match validated in final housing near target materials.
- Parser/diagnostics: include sequence & voltages; keep payload compact for fast reads.
- If considering PV/VEH, verify duty‑cycle math with real spectra (not lux) or real vibration PSDs—no assumptions.
Key Takeaways
- On‑demand energy beats ambient variability. Readers create predictable, short windows with enough energy for robust measure‑and‑report cycles.
- Sub‑1 GHz helps close the link. Lower path loss improves the harvested‑power margin at practical distances.
- Indoor PV and VEH are situational tools. Powerful under the right conditions, unreliable as a general baseline without environmental control.