Sensors and Markets

 March 21, 2012, SensorCon 2012, Santa Clara, CA—A panel of experts discussed the growing influence of sensors on the electronics market. Pallab Chatterjee moderated the panel. Panel members included Terry O’Shea from Intel, Kevin Shaw from Sensor Platforms, Janusz Bryzek from Fairchild Semiconductor, and MP Divakar from Microlytics.

Are there infrastructure limits to increases in MEMS?
O’Shea stated that nothing is an obvious block. We still need to work on reliability and analytics and better responses to inputs. Sensors are trustworthy and reliable in systems and communications systems with end-to-end data transfers are secure. Data storage is more than adequate and as secure as the sensors.
Shaw added that the mobile market is growing and moving well. Today, over 1 B phones are in use, and the number will grow to over 2 B in the next year, and most will have sensors. Augmented reality calls for a change in interfaces and standards are needed, especially for data transfers.
Bryzek demurred that standards are available for procedures after investments of over $1 B. the standards need resources and more organizations to work on the standards. Come companies are working on the issues, but the efforts are uneven and slow. One difference is that apps are increasing data traffic and loading the infrastructure and the people needed to analyze those data.
Divakar opined that the challenges are in the materials developments. Interoperability will come from standards and protocols. The back-haul infrastructure is a problem in the less developed countries.

Standards and infrastructure? This area is multi-disciplinary so new organizations are needed to cross the boundaries.
Shaw posited that Khronos has developed OpenGL, a screen input format with a common API. Major players are starting to move on the problems.
Bryzek offered growth will come from new industries in areas like sensor software analytics and data extraction. Some organizations like Semi, NIST, and others are looking into the issues. The standards need big company adoption.

IEEE 1451?
Bryzek said that since ’90, over 40 companies have participated in the standards work. Now the .5-.7 revisions are addressing test implementation and embedded signatures for interoperability and identification. Network sensor standards are evolving.

Prominence of data for location, calibration, etc., and metadata standards?
Divakar suggested TIA 50.1 does some of this
O’Shea appended IEEE 1173.03 covers the metadata, signatures, and calibration.

Lifecycle and infrastructure?
O’Shea offered the end user will define lifecycles through the value chain and the _ilties.
Bryzek mentioned that sensor networks on bridges will have to last decades.

Data standards are becoming business issues and the profit versus open issue is driving the industry towards standard processing for the sensors?
Bryzek offered lithography is the key. Silicon is cheaper than time. The next generation of sensors will be purely digital. Therefore, companies will have to change the sensors and functions to be capable for higher levels of integration. Commoditization causes business model changes.
The volume of data that networks of sensors can generate is an issue to be addressed. Intel instrumented a whole fab through tinyOS to monitor their equipment with accelerometers to check for excessive vibration. Efforts like this will case business model evolution, because software is not dependent upon silicon.
Shaw added that a billion units gets interesting. Even at a very low price, any growth still pays for itself. The infrastructure upgrades will be difficult, but the machine learning and big data analytics may help.
Bryzek considered that the medical industry will need codes for insurance so consumers can buy the monitoring equipment. A large market is for location and activity ads. The sensors and software detect the activities, and the advertisers will pay for that data. The increased performance in the sensors over time will allow for new markets. Historically, sensors have moved form defense-level performance to consumer, but the consumer products keep getting better. A loosely coupled sensor net and higher performance devices can monitor all types of functions at levels approaching defense levels.
Divakar suggested that volume will force manufacturers to change. For example, Sprint has a data pipeline that is not currently monetized effectively. The machine-to-machine data communications will overload the normal infrastructure, so Sprint could benefit from their dark fibers.

Privacy?
Bryzek considered concerns are segmented by age. The issue is how to incentivise users to give up their data. The big issue is the increase in apps and the growing permissions requirements. The next generation high performance systems will have features like face recognition and will sense emotions through changes in physiological measurements of heart rate, skin temperature, etc. The issue is when these data are forwarded to others.
O’Shea noted that the financial, and health areas are the most extreme areas for privacy. Unfortunately, the greatest number of data loss events is due to missing tapes. So far, there have been no major attacks on these databases.
Shaw agreed that sharing data has to be for something for you. You are the product and have to protect your own brand identity. One solution is to have some idea where the data goes. The phone becomes your intelligent agent and provides some local filtering of your personal data.

Privacy, ethics, and racial profiling?
O’Shea noted that the human genome testing raises ethical questions like could an insurance company drop you for a possible disease marker. The local culture and ethics also matter. Some of the issues are geographically specific.
Bryzek offered that the abuses will force the government to develop new policies and privacy will become a vehicle for more intrusive government into private affairs.
Shaw observed that Google scans all the gmails. In Japan, cell phone pedometer information is shared across networks to allow people to be competitive about their walking.

Policy directions for sensors in mobile devices? International issues?
O”Shea noted tht cell phone maps are disabled in Ireland.
Bryzek appended that local policies are embedded in the phone systems.

Hype versus the third revolution?
Bryzek stated that China and India had many workers. When the industrial revolution started, the new technologies increased individual productivity and displaces the workers. The second revolution started with the transistor and made knowledge more valuable than technology. Now, the third revolution is challenging whether information provides value. Since GDP is proportional to the number of people, expect the Chinese and Indians to become dominate again.
Divakar commented that more sensors, especially in medicine, bring coverage to more people and bring down the costs.
Shaw observed that the smart networks are not wise. No one knows what to do with all the data.

Future trends not in phones and health?
O’Shea offered temporal information and resolution. Sensors could detect a 10 angstrom change in displacement.
Shaw considered short term to have indoor navigation. Longer term will see gas phase chemical monitors bio-sensors for air bourn contaminants, and health safety.
Bryzek suggested smart grid monitors and the internet of things will enable condition-based maintenance. Weather monitors and other apps will require innovation.
Divakar postulated civil infrastructure will take over. We will not see big changes in consumer products.