February 24, 2012

Designs for Engineering Sensors

Date: Wednesday, March 7, 2012
Time: 10:00 - 12:15 pm
Room: E38-300 (292 Main St., Cambridge)

Open to the public.
RSVP appreciated, to Gayle Sherman, gsherman@mit.edu

Click here for the program. Each speaker presents their scientific findings on design challenges to engineering sensors, then answers questions from the audience. Please join us at 10:00 AM in the E38-300 conference room to listen, learn and participate in this stakeholder forum.


Michael Triantafyllou, MIT Dept. of Mechanical Engineering
Pressure Sensor Arrays Emulating the Lateral Line of Fish

Today, marine vehicles do not sense the oncoming flow or their own flow patterns. Substantial energy expenditures are required to keep a streamlined object at an angle to a current, while propulsion and maneuvering are seriously degraded in the absence of such flow information. To address this central issue in marine technology, we are developing the capability to save substantial amounts of propulsive and maneuvering energy, through sensing the vehicle’s own flow patterns and the oncoming flow, and employing effective control algorithms to adjust their attitude and position.

The significance of this capability can be understood when studying live fish, which are equipped with a lateral line, an array of pressure and velocity sensors which stretches along their side, but also appears at various points along their heads, where separation may occur. This capability allows fish to achieve substantial energy savings, and form effective schools. Using the lateral line, fish outperform by a wide margin similarly sized engineered vehicles with respect to energy efficiency, agility and maneuverability, since they can control the flow. Compare, for example, the salmon that travels over 1,000 miles upstream rivers without ever feeding, once they enter fresh water, with state-of-the-art AUVs; and yet the power level in fish is comparable to the power level available to man-made vehicles.

In the talk we will provide an outline of the developments in lateral line-like sensing, including applications in experiments and in the field.

Jeffrey Lang, Research Laboratory of Electronics at MIT
Flexible Underwater Pressure Sensors for Passive AUV Navigation

This presentation describes the development of a mechanically-flexible elastomer-based underwater pressure sensor array. When mounted on the hull of an aquatic vehicle, the array enables obstacle detection, identification and tracking, and can help reduce hydrodynamic vehicle drag by providing information about the surrounding flow. The pressure sensor array employs a piezoresistive carbon-black-PDMS elastomer as a strain-sensitive material, and uses individual diaphragms spaced 15 mm apart to convert pressure to strain. Each sensor in the array exhibits a 0.0014 fractional resistance change per 100 Pa, which yields a maximum pressure resolution of 1.5 Pa with a power consumption of 10 uW per sensor, excluding power dissipated by external amplifiers. No measurable cross talk between sensors has been observed. Each sensor can transduce up to a 1 kPa pressure differential across its diaphragm, though the sensor array can handle greater background pressures through a pressure equilibration scheme. An upper bound for the bandwidth of the diaphragms is calculated to be 940 Hz, ignoring viscous damping due to air, water, and the PDMS. Sensor operation while mounted to a hull with a 0.5 m radius of curvature has been demonstrated.

Franz Hover, MIT Dept. of Mechanical Engineering
Small Hovering Autonomous Vehicle

The inspection of underwater structures is a pressing need for ship and port security, dam maintenance, and monitoring of offshore platforms and equipment. In shallow water, human divers are undesirable not only from a safety and cost point of view, but also because the data product obtained by an automated procedure can be superior. To this end, we developed with our industry partner, Bluefin Robotics, a unique small hovering autonomous vehicle, the HAUV. The vehicle is built around two major sensor packages, a steerable Doppler velocity logger (DVL) - enabling hull-relative odometry - and an imaging sonar. The sonar combined with the DVL enables feature-based mapping and navigation, along with quantitative 3D mesh-building. In the stern area of a ship, inspection of the props and running gear pose a difficult 3D planning problem that we have addressed successfully with sampling-based coverage algorithms. These capabilities have been demonstrated on a number of real ships. The HAUV originally developed at Sea Grant is now a commercial product at Bluefin Robotics, and the basis of their largest contract to date.

Douglas Hart, MIT Dept. of Mechanical Engineering
High Resolution 3D Sensing for Surface Inspection and Precision Navigation in Close Quarters

Sonar systems have proven to be invaluable for UUV operations. However, the performance of sonar degrades appreciably at close range making it ill-suited to applications such as detailed inspection of surfaces, precision positioning feedback relative to moving objects, and navigation in close quarters through complex structures. Presented is a unique 3D sensor based on Active Wavefront Imaging (AWI) capable of providing high-accuracy, high-resolution 3D positioning information and detailed 3D surface mapping for inspection and navigation.

The MIT Sea Grant Stakeholder Forums provide an opportunity for our funded researchers to present their research and current findings to their peers and the public. Audience members and other presenting researchers are encouraged to ask questions and engage in dialogue. The goal of the series is to promote peer-to-peer networking, to connect research to those who can benefit from and apply the findings of MIT Sea Grant funded research, and for MIT Sea Grant to receive input on outreach programs whose goal is to deliver economic, social, and environmental benefits.


Sketch of a hovering AUV prototype, courtesy of Prof. Franz Hover

Quick Links