John A. Stankovic and Cyber-Physical Systems

John A. Stankovic is a professor of Computer Science at the University of Virginia. He has done a lot of research and work in areas related to Cyber-Physical Systems, the Internet of Things, and Wireless Sensor Networks. Stankovic has published many papers and has won eight Best Paper awards with two runner-ups. He is also the director of Link Lab, a department at the University of Virginia that does research into how technology can merge the digital and physical worlds.

Link Lab: A Look to the Future

Stankovic is the director the of Link Lab at the University of Virginia. Their four fields of research are: Autonomous Systems, Hardware for Internet of Things, Smart and Connected Health, and Smart Cities. The focus of this lab is to automate previously human done tasks and make them more efficient. A couple areas of study include driverless cars, smart storm water management, and improving the work of doctors and health professionals. A lot of Stankovic’s research here relates to Smart and Connected Health, which looks to integrate healthcare services into smart devices such as smartphones and smartwatches.

A few examples of his work relating to smartwatches includes MedRem, a medication reminder and tracking system, and Harmony which detects that the user is properly washing their hands. All of these are examples of cyber-physical system which is another topic that Stankovic is deeply involved in.

What are Cyber-Physical Systems?

A cyber-physical system is one in which computer system interacts with or monitors its surroundings and acts accordingly with the data that it gathers. This is similar to the idea behind the Internet of Things, in which sensors allow information to be gathered and exchanged over a network to potentially create a response or causing some action to occur. The two systems consist of the same underlying architecture and are very similar in nature. An example of a cyber-physical system is a collection of sensors monitoring temperature and humidity in an area and relaying that information to a central database for weather research.

Stankovic’s research of cyber-physical systems are primarily in the health and medical fields. The aim is to use available technology to gather information about the daily lives of patients and use that information to help both the patient and doctors. Examples of Stankovic’s work in this area include AsthmaGuide and KinVocal.

AsthmaGuide

AsthmaGuide is a smartphone application that collects data about asthma-afflicted patients and displays the information on a cloud network for both the patient and doctor to see. This aims to give doctors more complete information about a patient’s environment and what factors may cause their asthma. The idea is that since asthma can be triggered by various environmental factors such as temperature or humidity which differ between patients, it is hard for doctors to track exactly what may be causing problems for each individual patient. Since patients can only visit a doctor every so often, doctors lack information about the day-to-day lives of their patients which leads to incomplete data and hinders their diagnosis and treatment.

AsthmaGuide attempts to remedy these issues by using a smartphone to collect physiological and environmental information about the patient. This information is then shared with the doctors so they can better understand the patient’s situation and give better treatment.

The AsthmaGuide system can be broken up into four components:

1. A suite of sensors to detect environmental information as well as monitor patient physiological information.
2. A smartphone which is used as a hub interface for the entire system.
3. A cloud-based web application which stores and analyzes the data collected.
4. An advice suite which provides advice to the patient.

The sensor suite consists of an electronic Bluetooth stethoscope, spirometer, and pulse oximeter to gather information about a patient’s lung sounds, blood oxygen levels, and other data points. A Sensor-drone is used to collect information about the environment such as temperature, humidity, and gases such as oxygen, carbon monoxide, and ozone.

After the data is collected, a smartphone app is used to aggregate and upload the data to the cloud-based web application. Of note here is that the collection of some data such as lung sounds are based on manual input. Patients and medical officials can view the collected data on the cloud web application. There are sorting features here in which a user can view readings from certain points in time or check for certain environmental factors. Finally, the advice infrastructure will give the patient advice on what to do. However, this advice is very general as they do not wish to endanger the patient. The system can also give a warning or alert of certain factors are detected such as a high pollen count.

Overall, AsthmaGuide showcases a cyber-physical system which looks to enhance the capabilities of medical professionals by providing them a wealth of information that they never had easy access to previously. By combining sensors with a ubiquitous device like a smartphone and cloud technologies, this information can be readily gathered and used to improve the aid that doctors can provide to their patients.

KinVocal

Along the same lines as AsthmaGuide, KinVocal looks to detect agitation of patients with dementia. This includes asking for help, screaming, crying, repetitive sentences and more. The goal is to help physicians provide better diagnosis and care for their dementia patients.

The system looks to accurately detect and record all eight vocal agitation events that are listed in the Cohen-Mansfield agitation inventory. The eight are as follows:

• constant unwarranted request for help
• making sexual advances (verbal)
• crying
• screaming
• laughing
• cursing
• speaking in repetitive sentences
• negativism

The KinVocal system uses a Microsoft Kinect sensor to collect sound information. The sensor also serves to handle some noise elimination. The information is then processed into text and analyzed by different modules. These modules look to detect different actions listed above. It uses a combination of the text and sound inputs to accurately detect and categorize each of the eight vocal agitation events.

So far with testing the system on movies, YouTube videos, small home deployments, and controlled experiments, the system has achieved a detection accuracy of nearly 90% for all categories with some categories such as laughing, crying, and screaming being accurately detected with over 96% accuracy.

KinVocal is another example of a cyber-physical system. By incorporating a Microsoft Kinect sensor, it can detect agitation and accurately categorize it leading to better diagnosis for physicians. Using technology, more data can be obtained than was previously possible. This data could lead to better treatments for the future.

Conclusion

As you can see, AsthmaGuide and KinVocal have many similarities — they both look to help medical professionals provide better service by giving them access to more information about their patients. These two works are just a small example of how cyber-physical systems can benefit the world.

John A. Stankovic has contributed a ton to the field of cyber-physical systems and incorporating technology to aid medical professionals in their work. His work on AsthmaGuide and KinVocal serve as examples of what can be done by combining technology with the medical field. These works can form the basis and serve as inspiration for future research and work in not just the medical field but many more fields such as classrooms, environmental observations, and more.

Other Works

Stankovic also has done a myriad of other work such as research on Wireless Sensor Networks, which look to develop solutions to building systems in smart homes and smart cities. He has written about using technology to make smart cities a safer place with his CityGuard project. Here is a link to his University of Virginia faculty page if you are interested in more of his works.