Abstract Remotely detecting the physiological state of humans is increasingly important for rehabilitative robotics (RR) and socially assistive robotics (SAR) because it makes robots better-suited to work more closely and more cooperatively with humans. This research delivers a new non-contact technique for detecting heart rate using a high precision, single-point infrared sensor. The proposed approach is an important potential improvement over existing methods because it is capable of collecting heart rate information unencumbered by biofeedback sensors, complex computational processing or high cost equipment. We use a thermal infrared sensor to capture subtle changes in the sub-nasal skin surface temperature to monitor cardiac pulse. Temperatures are obtained by tracking the sub-nasal region of the face, continuously targeting and sampling the infrared sensor. This study extends our previous research in which breathing rate is automatically extracted using the same hardware. Experiments conducted to test the accuracy of the proposed heart rate detection system show that in 72.7 percent of typical cases heart rate was successfully detected within 0-9 beats per minute as measured by root-mean-square error.