MAGIC Smart Home Project

The MAGIC Smart Home Project has been made possible in part through the Community Care Challenge grant from Point Breeze Credit Union.  

Executive Summary

The MAGIC Smart Home Project is an innovative collaboration demonstrating a variety of broadband enabled technologies in a residential setting that will gather a wide variety of data, provide intelligent monitoring, analysis, and alerting to make a measurable positive impact on the quality of life of residents. Specifically, the project will combine telemedicine and in-home monitoring and controls to decrease Emergency Department visits, decrease unplanned medical appointments, and decrease staff time spent on unplanned medical care for a population of medically fragile adults with intellectual disabilities.

Scope

Using the community wide gigabit Westminster Fiber Network (WFN), the partners will connect two residential homes managed by a local non-profit providing residential services to adults with intellectual disabilities (Target Community and Educational Services). Inside each home, a variety of sensors will be incrementally deployed and connected to an intelligent integration platform, which will collect data, analyze, report, and alert prospectively and in real time, as well as provide administrative support to staff, and perform typical smart home functions with mechanical system and appliance management. The open architecture of the integration platform allows for unique and flexible combinations of sensors to solve specific problems for residents and staff in the homes. The integration platform connects to a variety of authorized users via a Software Defined Networking (SDN) solution which will create customized, highly secure, ad hoc network connections between locations only when and how needed. Together, the integration of these solutions will create the fundamental operational unit of a truly Smart Community.

Use Case

Unlike most prior smart home demonstrations, the MAGIC Smart Home Project will combine “traditional” smart home technologies (environmental controls and monitors) with telemedicine capabilities (physiologic monitoring and remote interaction) to test and validate several important new functionalities:  

  1. Use of Software Defined Network (SDN) provided by EntryPoint to enable highly secure and reliable situational routing of information only to interested and authorized users, only when required; 
  2. Integrating multiple sensor modalities to create additional levels of problem detection and monitoring, coupled with intelligent integration to enable a higher level of care in the home currently impossible with existing web based interfaces over conventional broadband; 
  3. Data collection and analysis to enable more efficient care and demonstration of specific cost-savings and outcome improvements in the home setting; 
  4. High profile demonstration of the security, privacy and reliability of new network capabilities and sensing modalities that preserves the dignity and privacy of residents, overcoming concerns about intrusive technology, while at the same time solving real problems and exceeding the most stringent requirements for data security, far in excess of what is currently possible with web-based interfaces;  
  5. Implementing these improvements in a cost-effective, scalable manner with tangible, measurable savings and cost-avoidance in excess of implementation expense.

Impact Measurement

Key Performance Indicators will initially include:

  • ED visits
  • Urgent Care visits
  • Hospital admissions
  • Unscheduled doctor appointments
  • Exacerbations of underlying chronic diseases
  • Hours of staff time spent supervising all of the above, quantified weekly.

Baseline measures will be collected over the prior two year period from existing documentation maintained by Target Community and Educational Services, and using a prospective analysis plan, real world data elements will be collected. To the greatest extent possible, the project will use pre-defined data elements, a common definitional framework, and pre-specified time intervals for data element collection and outcome analyses. In addition, a data quality assurance plan will evaluate data collection procedures and integrity.

Specific Scenarios

Current solutions either focus on in home control managed by the resident (“lights off;  temperature 72 degrees”), or the aggregation of sensor data brought back to a common monitoring platform, overseen by a central authority, and may be shared with other authorized users for monitoring only (family members).

The MAGIC Smart Home Project will demonstrate a level of in home intelligence that will enable the following types of capabilities:

  1. After 10 pm, door sensors are armed. Later, motion detectors in the front yard activate lights and outdoor camera and capture a stranger in the front yard. At 11:33 pm, the front door sensor alarms. The front door camera is activated, and a person is visualized. The home interface automatically provisions a secure connection through SDN, and sends an alert to the hand held device of the home manager, as well as the alarm company, with a video feed from that camera. Both the home manager and the alarm company activate a call to 911. Until the sensor is activated, the camera is dormant, and no network connection exists, preventing unauthorized use of the camera. After the 911 call, the alarm company sends the same video feed, now from all cameras in the house, to the responding police officer.
  2. Same scenario, but upon entry, the house detects the ID badge of one of the staff, who is coming home late from a day off. All alarms are deactivated, staff remain undisturbed.
  3. A client is watching TV and falls asleep. The house monitoring system detects the client on the couch in front of the TV without movement for two hours and alerts the home manager, who checks on the client to find him asleep in the couch.
  4. A client gets out of bed and falls in the middle of the night. The house monitoring system detects no movement and the client not in bed, alerts the home manager after 2 minutes, who finds the client having a seizure. Manager alerts 911 with her handheld device.
  5. A client with a history of diabetes and hypertension is having complications. Daily vital signs are automatically and securely uploaded across a temporary network connection to the primary care physician (PMD). Serial measurements record high pressures, and a call to the PMD initiates a video consult. A secure connection across the gigabit network using SDN enables a 4K UHD video connection and the PMD converses with and visually examines the client, while reviewing all vital sign data collected in the residence. The PMD consults in real time with the Emergency Department, and together they determine the client does not need to visit the ED and the PMD makes a medication adjustment. All transmissions of patient data and video occur over a completely HIPAA compliant, temporary, private network that never touches the Internet.
  6. The same client has a diabetic ulcer which requires daily wound care. A wound care specialist can see the wound and give advice about treatment using a 4K UHD video connection, saving the client and staff a trip out of the home each day for follow up.
  7. A client is having a behavior problem. Staff implement restraint protocols. For quality assurance monitoring, the home manager activates the in home camera only in the room where the client and staff are present. The client is released from restraint per protocol, but continues to act out, so is left alone to deescalate. The home manager monitors the client remotely through the in home network to her handheld device, and alerts her supervisor, activing a temporary secure video feed allowing the supervisor to remotely see the video feed.  At the same time, the home manager alerts the family member POC, who can also monitor the video feed in real time.  Later, the client is found to have bruises from the tantrum. During a visit with the PMD, concerns are raised about the nature of the injuries, but after reviewing the video, and discussing with family members who also reviewed the video, all concerns are addressed and resolved.
  8. Staff dictate daily care notes into the records system via an interface in the home with speech recognition. All notes are logged into the electronic record and are immediately available to the next shift, managers, and state auditors when needed during regular reviews.

Project Timeline

Phase 1- installation and lighting of fiber connections to demo home locations (in process, expected completion Feb. 2017)

Phase 2- installation of intrusion detection (in process), integration platform, and SDN to initial locations. (Jan.-Feb. 2017). Organization of baseline KPI data from existing Target Community and Educational Services records. Testing and validation of integration and interoperability. Connect to hospital.

Phase 3- implementation of efficiencies and process improvements for in-house staff using integration platform. (2017)

Phase 4- implementation of initial biometric and physiologic sensors for collection of clinically relevant and useful data for medical management. (2017)

Phase 5- expansion of networked parties to include local hospital, care management resources, and other interested parties.

Phase 6- expansion to other communities.

Partners/Interested Parties