CHARLES HOWARD SOBEY — Department of Energy SBIR Phase II: C52-05d

Description

There is new value in old data. “Legacy” data sets are used to train machine learning models, providing new insights and increasing, for example, the range of climate model predictions. Decades-old oil and gas seismic surveys, on legacy 9-track tapes, are being re-evaluated for clean energy exploration. They can reveal lithium deposits for energy-dense batteries, sources of geothermal energy, and locations for carbon sequestration. Troves of other types of legacy data are stored unexamined, offline, delivering no value. Worse, they incur storage and maintenance costs, are deteriorating, and are stored in obsolete formats. Large data recovery companies can retrieve data from legacy tapes and migrate the data to the cloud for analysis. The best companies maintain vintage tape drives and have technicians and engineers skilled in handling, cleaning, and repairing damaged media. But recovering damaged media is costly – when even possible. Old tapes suffer from sticky-shed syndrome or vinegar syndrome, which makes simply reading a tape destructive, losing data forever. The hardware solution developed under this award senses a tape’s magnetic data without head-tape contact, reading over damaged areas and eliminating head wear. With Phase I funding, an experimental tape transport system supporting noncontact head positioning was developed. A custom, low-noise amplifier was designed and built for using abundant, modern magnetic recording heads to sense legacy 9-track tapes. Analysis of novel noncontact readback signals is underway. Phase II funding will back development of a transport system with precise speed and tension control, supporting experimental noncontact positioning systems. Servo-controlled positioning will be developed along with automatic calibration algorithms. A software-controlled 2nd generation amplifier, with more channels, will be built to enable higher-density heads to provide finer sensing of tape surfaces. Head fields will be characterized to enhance noncontact signal resolution. These developments will support better reading of damaged 9-track tapes and enable analysis of legacy data cartridges and video tapes. Legacy magnetic tapes contain irreproducible scientific data, including for the Department of Energy’s National Nuclear Security Administration mission. Dual-use of this technology will benefit oil and gas surveys; recordings of classic performances; financial, medical, and demographic records; forensics; telemetry data; field recordings of vanishing languages; metadata scholarship; and digital heritage/preservation efforts. This noncontact magnetic sensing technology may lead to low-cost magnetic microscopy, electrical activity sensing (possibly for non-invasive brain studies), metal fracture detection, and DNA analysis.