Human Computer Resource Perception and the Swimming Pleasure Measure

GBF2 proposes a new kind of situated environmental descriptor and a numerical procedure by which to produce it. The Human Computer Resource Percept (HRCP) is an environmental descriptor that combines human intuition expressed in language and sensor data processed by computers.

The HCRP in GBF2 is coined the swimming pleasure measure (SPM) and covers three different modes of knowing. First, it includes human perception (human resource percepts (HRP)). Human beings are formidable sensors. Skin, eyes, nose, mouth and mind are able to detect and reason based upon subtle environmental cues. Second, it includes the best sensing technology available and in use by environmental monitoring agencies. Computer Resource Percepts (CRP) such as sensors coupled with machines can know things people can not. Third, the SPM includes subjective preferences, opinion and intuition.

The HCRP's human input is a crowd-sourced summary of the current beach visitors' perceptions of the water (based on smell, taste and visual appeal). For recreational waters, the commonly accepted quality indicator is the presence or absence of fecal contaminants. The most common microbial indicator of pathogens in fecal matter is E.coli. The inability of E.coli to grow in water combined with its short survival time in water environments means that the detection of E. coli in a water system is a good indicator of recent fecal contamination. For this reason, GBF2 uses the presence of E.coli (and other measures people cannot detect such as pH) in its evaluation algorithm as a corrective factor that can override human opinion. The SPM follows the mean of current beach visitors' opinions as long as these independent parameters are below acceptable (here EPA) risk levels. Otherwise the SPM follows the results from the sensors and contradicts people's perception of the beach. The SPM and the HRCP it is based upon are a new kind of metric of beach water quality, offering a combination of situated subjective perception and sensor evaluation.

Environmental monitoring is both science and social practice. Rules can be arbitrary, enforcement selective and risk thresholds variable. Including the opinions of people as social actors with particular interests acknowledges the significant contribution of this factor. Including state of the art sensor measurements ensures that that these viewpoints are held in check by the other data. Combining them numerically creates, if nothing else, a basis for discussion of a formal framework of a new class of situated environmental descriptors.

The following equations describe the relationship between the HCRP, HRP, CRP and the SPM:

hcrp = hrp * π(crp(i)) SPMeffective = f(spm_w)*π((E.coli, pH, ...))

The scale of the SPM (1 to 10) is chosen based on the ability of people of all linguistic and cultural backgrounds to relate experiences of the body to scales of the body. The hand is a particularly potent scale system, offering a five and ten count on demand. This links GBF2's algorithm to pre-historic origins of number and ordered systems that retain communicative potential even in the age of ubiquitous informatics.

The rationale for this generalized approach is the growing need for cross-cultural and global resource monitoring metrics that can operate in and adapt to different parts of the world. Water is a globally shared resource; protecting it locally for the benefit of few is increasingly irrelevant.

Data sources and data types

GBF2 collects many kinds of data from different sources. GBF2 has an onboard weather station to track weather features and a submersible sonde that measures many parameters typically used in water quality evaluation. GBF2 is connected to an on-site bio incubator that evaluates e.coli and total coliform concentrations. GBF2 also has 'data-guards' who interview swimmers just after they exit the water to inquire about their swimming experiences.

This table summarizes the various data GBF2 considers:

YSI 6600V2 water quality data sonde
dissolved oxygen[mg/L]how much free oxygen is in the water?
turbidity[NTU]how clear is the water?
chlorophyll[μg/L]how much algae are present?
pH[H+]how acidic or basic is the water?
water temperature[C]how warm or cold is the water?
salinity[ppt]how high is the ionic content of the water ? / how salty is the water?
total dissolved solids[mg/L]how high is the combined content of inorganic and organic substances in the water?
LaCross weather station
air temperature[C]how warm or cold is the air?
wind speed[m/s]how strong are the winds?
relative humidity[%]how humid is it?
dew point[C]temperature at which the air can no longer hold all of its water vapor (always less than the air temp)
barometric pressure[mmHg]how high or low is the atmospheric air pressure?
rainfall (1hr, 24 hr)[mm]how much rain has fallen in the last hour or over the last day?
forecast[n/a]what is the weather forecast for the next few hours?
Endetec B16 Microbial Detection system
e-coli[colonies/mL]how much E.coli contamination (possibly from sewer run-off) is in the water?
total-coliform[colonies/mL]what is the sum of all forms of coliforms in the water?
Cetacean Research CR1 hydrophone
underwater sounds[n/a]what kind of sounds are there underwater?
Hummingbird 788ci sonar and gps system
gps location[lat, long]what are GBF2's coordinates?
wave height approximation[m]how high are the waves where the buoy is located (approximation)?
water depth[m]how deep is the water where the buoy is located?
cloud cover[%]what percentage of the sky is covered with clouds?
cleanliness of the beach[n/a]how clean is the beach at the beginning of the day?
number of people at the beach[n/a]approximately how many people are at the beach?
random observations and comments[n/a]comments outside of categories above
experience of the water[1-10]subjective, personal experience of the water, mapped 1-10
experience of the beach[1-10]subjective, personal experience of all aspects of the beach, mapped 1-10
swimming ability[n/a]self-defined swimming ability, non-swimmer to expert
frequency of visits[n/a]how often an interviewee visits this particular beach
duration of visit[hrs]how long an interviewee intends to stay at the beach
comments[text]all kinds of comments welcome
observational data[age bracket]age bracket of interviewee (child, teenager, adult, elderly)
free form interview[n/a]open-ended mic; no restrictions

Data collection methods and technology

Sensor data is collected throughout the day. Readings from a nearby NOAA weather station are collected every 60 minutes. Readings from the on-board weather station are made every 15 minutes. Readings from the submersed water quality sonde are collected every 15 minutes. Short anonymous interviews are conducted about 20 times per day within a cross section of the beach visitors. Each person is asked six questions, including one direct request to describe the perceived quality of the water on a scale of 1 to 10, where 1 is the lowest and 10 the highest possible mark. Additionally we perform in-depth audio recorded interviews with select beach visitors interested in sharing their stories about their relationship to the beach and water in general.

GBF2's software is written in python and php for linux.

team publications presentations contact