SBIR-STTR Award

In addition to their ubiquitous abilities for navigation at small spatial scales, a host of animals migrate long distances through the open ocean or skies. The mechanisms underlying these powerful navigational abilities are slowly being decoded. This progress offers the opportunity to develop low cost navigational systems for use in future autonomous underwater vehicle (AUV) designs. As future AUVs come to emphasize low size and cost for deployment of larger numbers of AUVs for a given task, such systems are desirable. In this proposal, we focus on low cost, bio-inspired multisensory navigation through the combination of three different sensory modalities: active electrosense, geomagnetic sensing, and inductive magnetosense. We propose to use geomagnetic sensing for sensing the local direction of the magnetic field, inductive magnetosense for velocity sensing, and active electrosense for high resolution localization and object collision avoidance.

Benefit:
Our research suggests that our approach offers the following advantages relevant to Navy missions: Stealth: These sensors do not require light to operate. The active modality, active electrosense, occurs at such weak signal strengths as to be below thermal noise a few vehicle lengths away. Efficiency: Two of the three sensors we will use are passive, and thus have minimal power requirements. Active electrosense functions at very low power. See-through capability: Active electrosense can peer through opaque surfaces to detect buried objects and characterize their impedance, such as detection of buried metallic remnants associated with sunken ships or lost airplanes. Detection of live objects (such as for live swimmer tracking) is also possible due to the higher capacitance of live tissue. Multiscale resolution: Geomagnetic sensing gives us longitude and latitude with low resolution, with higher resolution where there are mapped geomagnetic anomalies. At the other extreme of length scales, active electrosense gives us precise (to within 0.5 mm: Solberg et. al 2008) localization given a preexisting map. Bridging these large and small spatial scales, inductive magnetosense will be used to estimate velocity for dead reckoning of position at midrange spatial scales.

Keywords:
magnetoelectrosensory navigation system, magnetoelectrosensory navigation system, geomagnetic sense, magnetosense, bioinspired sensing, active electrosense, multisensory navigation

Many aquatic creatures transduce magnetic and electric fields for navigation cues and feedback control. Such sensing capabilities would greatly advance AUV technology. We will continue the co-design of sensing and motion capabilities along with empowering algorithms for underwater perception and navigation. We have made significant advancements in active electrosense and inductive-magnetosense in Phase I, and we will continue to improve its capabilities by employing low-noise digital electronics. We will also investigate the imaging of nearby spatial electrical impedance using a new architecture of active electrosense. We will develop new algorithms for exploration, feature classification, and navigation without a sensor map. The new hardware and algorithms will be demonstrated on a 6 degree-of-freedom fully-automated gantry. Stereotyped motions emerging from the exploration algorithms in the gantry test bed will give insight to desirable motion capabilities for an AUV with electrosense. Evolution suggests that the propulsion system of the weakly electric fish be an excellent complement to electrosense. Thus, we will build and test a robot with a ventral ribbon fin and pectoral dive planessimilar to the propulsion system of weakly electric fishwith an integrated electrosensory system for closed-loop control.

Benefit:
Our research suggests that our approach offers the following advantages relevant to Navy missions: Stealth: These sensors do not require light to operate. The active modality, active electrosense, occurs at such weak signal strengths as to be below thermal noise a few vehicle lengths away. Efficiency: Two of the three sensors we will use are passive, and thus have minimal power requirements. Active electrosense functions at very low power. See-through capability: Active electrosense can peer through opaque surfaces to detect buried objects and characterize their impedance, such as detection of buried metallic remnants associated with sunken ships or lost airplanes. Detection of live objects (such as for live swimmer tracking) is also possible due to the higher capacitance of live tissue. Multiscale resolution: Geomagnetic sensing gives us longitude and latitude with low resolution, with higher resolution where there are mapped geomagnetic anomalies. At the other extreme of length scales, active electrosense gives us precise (to within 0.5 mm: Solberg et. al 2008) localization given a preexisting map. Bridging these large and small spatial scales, inductive magnetosense will be used to estimate velocity for dead reckoning of position at midrange spatial scales.

Keywords:
bioinspired sensing, geomagnetic sense, magnetoelectrosensory navigation system, magnetosense, multisensory navigation, active electrosense