Some systems to measure voided volume and ow rate have been developed. Clinically, urination parameters are measured using uro owmeters that have special receivers such as cups or bowls. Since these uro owmeters were developed for clinical use, home use is dif cult and inconvenient. Many of these devices require equipment cleaning; additionally, most are too expensive for home use. To address these problems, we developed a method to measure voided volume by noncontact matrix temperature sensors that are installed under a toilet seat. The basic concept is as follows. Urine is excreted at core temperature of 37 C. The heat radiated from urine during excretion is measured by noncontact matrix temperature sensors, and the measured radiated heat is converted to urination volume. A preliminary study was conducted to estimate the voided volume using an actual toilet bowl. The noncontact matrix temperature sensors simultaneously measure the temperature of falling water at 37 C in 16 areas using a matrix of four lines and four rows. Four noncontact matrix temperature sensors were installed at the front, rear, left and right of the underside of a toilet seat. The toilet seat equipped with the four sensors was installed on a toilet bowl. The position of the falling water was xed at 120 mm from the front sensor. Water volumes of 100, 200 and 300 ml were passed vertically at ow rates of 10, 20, 30 and 40 ml/s. As a result, the surface of the toilet bowl was slightly heated from the heat of the falling water, and the toilet bowl retained the heat after the water had fallen. To eliminate overestimation of heat from the toilet bowl, we proposed two analytical methods: a bias temperature elimination method and a time limitation method. For all four ow rates, the variation of U with a proportional volume obtained by the time limitation method was smaller than that by the bias temperature elimination method.

We developed an excretion monitoring system equipped with non-contact temperature sensor and a semiconductor gas sensor installed under a toilet seat. The gas sensor is able to detect odors due to human excretion by transiently increasing the output. The peak amplitude and duration of the gas sensor output due to defecation were signi cantly greater than those due to urination. These results demonstrate the capability of the gas sensor in distinguishing whether the excretion is urination or defecation.