Tuesday, May 29, 2012

Softbank 107SH Pantone 5 Geiger counter phone: Tool or toy?

[UPDATE: Based on the link provided below, this phone does not (apparently) incorporate a standard GM tube, so it may be more rugged than I thought originally. However, voltages still have to be reliably converted to meaningful output over the life of the phone. I still believe this will be nothing more than a gimmick, but perhaps for different reasons that I originally thought.]

Providing accurate radiation data to the public is extremely important. A larger database that includes data of substandard quality is not better than a smaller one. The decreased accuracy hinders our ability to properly detect the broader spatial and temporal trends. Therefore, I'm skeptical that it is appropriate to build a Geiger counter directly into a mobile phone, as has been done with the Sharp Pantone 5 (107SH) smartphone just announced by Softbank Mobile.

In the end, I imagine it will be no more than a gimmicky toy (the main contribution of which was increased handset sales). Here's why.

Raw counts must be converted to dose though calibration with a known source, typically Cs-137. Measurements of radiation emitted from any other radionuclide will be inherently less accurate. According to the manufacturer of the primary Geiger counter used by Safecast, the $700 Inspector Alert, this unit's dose conversions exhibit a +/- 15% accuracy, immediately subsequent to proper calibration. Over time, readings produced by any analytical equipment drift, causing reported values to become increasingly inaccurate. Therefore, the device must be periodically calibrated - preferably preemptively, as opposed to after conversions become unreliable.

Calibration frequency of any analytical equipment is dependent on application. For equipment used in public safety, local government regulations will require calibration at a set interval. When collecting data for scientific studies, analytical equipment must be calibrated prior to each usage. For quality control during use, calibration standards must be intermixed with experimental samples to detect and correct any instrumental drift.

Calibration would also be needed after any shock to the device, the kind that would be sustained by, for example, dropping a mobile phone.

Any other way of doing it is just playing around, and toys are for playing.

 300 cpm ≈ 1.0µSv/hr ≈ 1.1Bq/cm2
callibrated [sic] for Cs137 ± 15%
Image source.
Data collected by hobbyists in this manner is similar to a regular guy aiming to hit a barn door with a baseball from twenty paces - a major league pitcher is not needed to be accurate enough, and I'm sure the counters in the Pantone 5 can hit the barn door. Keeping up the baseball analogy, these (and Safecast's) data would be best considered in the context of the ballpark, as in you know where the ballpark is, but you can't say who's on first. Or on second. Or in centerfield. Just, that they're in the ballpark.

On to the phone.

The Softbank 107SH will be available from mid July. In addition to the standard WiFi, Bluetooth, and GPS, here are the major features:
  • Android 4.0 Ice Cream Sandwich
  • Water resistance (IPX5 and IPX7)
  • Geiger counter (0.05 - 9.99 µSv/hr range)
  • W-CDMA 900 MHz "Platinum Band", 1500 MHz, 2100 MHz
  • One Seg TV
  • Keitai Osaifu mobile payments
  • Infrared port
  • Earthquake warning mail (緊急速報メール) coming with a future update
  • 3.7 (854×480 pixel) display
  • 4GB ROM
  • 1GB RAM
  • 8 color choices


  1. I agree there is not so much interest in a low accuracy monitor. This is probably why the prototypes showed in CEATEC last year by Docomo never end-up to become real products.
    I'm interested to have your opinion on the sensor we produce: http://www.horiba.com/jp/process-environmental/products/environmental-radiation-monitor/details/pa-1100-16189/
    I cannot talk about the sensor itself as I was only involved in the android app development but I would welcome any suggestions about features that this kind of smartphone app should include.

  2. The detector seems to be silicon-based. Probably it is counting the photons, so its drift should be negligible. The objection that conversion from counts to dose strongly depends on the source is a much greater problem. A small Si detector is much more sensitive to x-rays from a CRT tv screen than to Cs-137 gamma radiation.

  3. Thanks for the comment, Pieter.

    The examples I was using regarding instrumental drift are probably less appropriate to a geiger counter used to detect ballpark radiation, rather than say a mass spectrometer designed to detect specific atomic masses over 7 orders of magnitude.

    Though I still think that the readings displayed on the screen will be affected over time, maybe not necessarily due to the tube.

  4. My first opinion is that your price point is too high by a factor of 3. I do like the idea of an external sensor, one that won't be subject to the abuse that a mobile phone is going to take. Pull it out when you need it, then put it away.

  5. Reuters announcement on this: http://www.reuters.com/article/2012/05/29/us-japan-phone-idUSBRE84S05I20120529

  6. The smartphone in the company's "Pantone" series will come in eight bright colors and include customized IC chips made by Sharp Corp that measure radiation levels in microsieverts per hour.

    So, this is not a GM tube, so it may be more rugged than I thought originally. However, voltages still have to be reliably converted to meaningful output over the life of the phone.