{"id":450,"date":"2021-02-01T07:45:17","date_gmt":"2021-02-01T07:45:17","guid":{"rendered":""},"modified":"2023-08-02T15:55:08","modified_gmt":"2023-08-02T15:55:08","slug":"research-2","status":"publish","type":"page","link":"https:\/\/robotsforaging.cs.unh.edu\/?page_id=450","title":{"rendered":"Research"},"content":{"rendered":"\t\t
\n\t\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t\n\t\t\t\t\t
\n\t\t\t

Overview<\/h2>\t\t<\/div>\n\t\t\n\t\t\t\t
<\/div>\n\t \n\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t

The MARSS framework has four evidence-supported modalities of AD care: activity engagement and assistance, telehealth, home safety, and caregiver-care recipient connectivity. We also have a training program to enable a non-technology expert (caregiver, family member or health professional) to scale and program the modalities to fit with the disease severity and context of the IAD.<\/p>

MARSS is being developed and tested in accordance with the NIH Stage Model of Intervention Development. Stage I and II involve scaling up the existing lab-based model of MARSS for a pilot test in the community (with 8 dyads of IAD and caregivers) and verify its implementation fidelity, robustness as well as behavior change techniques to optimize target engagement of IADs and caregivers (2022-2024). Stage III involves an 18-month randomized controlled trial to validate the real-world efficacy of MARSS (2025-2027). We will recruit 60 dyads in two staggered cohorts of IAD and caregivers and randomly assign them to the intervention (n=30) or a control group (n=30). We will gather repeated measures data on the IAD\u2019s functional independence, safety, and physical and cognitive health, and the caregiver\u2019s perceived care burden, autonomy and wellbeing over nine data points, 2 months apart. To account mechanism-focused change, we will objectively collect data on the technology\u2019s utilization to tease out the influence of the MARSS\u2019s modalities on the intervention outcomes.<\/p>\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t

\n\t\t\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\"Picture1\"\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\n\t\t
\n\t\t\t\n\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t

Technical Innovation\u200b<\/h3>\n\t\t\t\t\t
<\/div>\n\t\t\t\t<\/div>\n\t\t\n\t\t\t\t\t\t
\n\t\t\t
\n\t\t\t\tA large collection of care protocols, termed as artificially intelligent care protocol (AICP), realizes the MARSS intervention. An AICP is nothing but a sequence of care-related tasks planned by an artificially intelligent planner (AI planner) and to be executed by a mobile robot.As shown in the figure 1, MARSS hardware system consists of SH sensors that gather information to build awareness about different care needs, and a mobile robot that leverages its limited anthropomorphism and mobility to deliver care to IAD. We are using mobile platforms (such as Jackal) and quadrupeds (such as Go 1 EDU) as robots. A ROS-based layered software architecture, as shown in figure 2, drives MARSS hardware system. The Planner Layer has a planning executive and the AI planner \u2013 implemented through PlanSys2 \u2013 which connect and synchronize all sensor information, generate a planning problem on the fly based on pre-coded PDDL domain files, call the AI planner with the problem instance, and dispatch the plan. As it generates a plan, the AI planner is aware of the executable tasks available in the Task Layer. \t\t\t<\/div>\n\t\t<\/div>\n\t\t\n\t\t\t\t\t\t\t<\/div>\n\t\t<\/div>\n\n\t\t<\/a>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t\t\t
\n\t\t\t\t\t\t
\n\t\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\n\t\t
\n\t\t\t\n\t\t\t
\n\t\t\t\t\n\t\t\t\t\t\t
\n\t\t\t
\n\t\t\t\tTasks in the Task Layer are pre-coded through PDDL. A robot needs a set of basic skills to execute any task e.g., navigation, localization and those skills are the content of the Skill Layer. Skill Layer is implemented through integrating state-of-the-art robotics algorithms available through various ROS packages. The HIPAA-compliant security layer ensures that the robot system is resilient to network attacks (Privacy security) and robust to complete the required service (AI\/Software security). The privacy security module supports data encryption, access control, logging, backup, and firewall. A comprehensive penetration testing will be employed to evaluate and protect the safety of the network components. The software security module adopts fuzz testing, an advanced automated software testing technique that can rapidly search for states likely to cause system failures. The fuzzer runs for several rounds, where each round collects feedback from the previous round and aims at reaching deeper components that have not been tested before.\t\t\t<\/div>\n\t\t<\/div>\n\t\t\n\t\t\t\t\t\t\t<\/div>\n\t\t<\/div>\n\n\t\t<\/a>\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t
\n\t\t\t
\n\t\t\t\t\t\t\t\t
\n\t\t\t\t
\n\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\"\"\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t<\/div>\n\t\t\t\t<\/div>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t\t<\/div>\n\t\t<\/div>\n\t\t\t\t\t\t\t<\/div>\n\t\t<\/section>\n\t\t\t\t\t\t\t<\/div>\n\t\t","protected":false},"excerpt":{"rendered":"

Overview The MARSS framework has four evidence-supported modalities of AD care: activity engagement and assistance, telehealth, home safety, and caregiver-care recipient connectivity. We also have a training program to enable a non-technology expert (caregiver, family member or health professional) to scale and program the modalities to fit with the disease severity and context of the IAD. MARSS is being developed and tested in accordance with the NIH Stage Model of Intervention Development. Stage I and II involve scaling up the existing lab-based model of MARSS for a pilot test in the community (with 8 dyads of IAD and caregivers) and verify its implementation fidelity, robustness as well as behavior change techniques to optimize target engagement of IADs and caregivers (2022-2024). Stage III involves an 18-month randomized controlled trial to validate the real-world efficacy of MARSS (2025-2027). We will recruit 60 dyads in two staggered cohorts of IAD and caregivers and randomly assign them to the intervention (n=30) or a control group (n=30). We will gather repeated measures data on the IAD\u2019s functional independence, safety, and physical and cognitive health, and the caregiver\u2019s perceived care burden, autonomy and wellbeing over nine data points, 2 months apart. To account mechanism-focused change, we will objectively collect data on the technology\u2019s utilization to tease out the influence of the MARSS\u2019s modalities on the intervention outcomes. Technical Innovation\u200b A large collection of care protocols, termed as artificially intelligent care protocol (AICP), realizes the MARSS intervention. An AICP is nothing but a sequence of care-related tasks planned by an artificially intelligent planner (AI planner) and to be executed by a mobile robot.As shown in the figure 1, MARSS hardware system consists of SH sensors that gather information to build awareness about different care needs, and a mobile robot that leverages its limited anthropomorphism and mobility to deliver care to IAD. We are using mobile platforms (such as Jackal) and quadrupeds (such as Go 1 EDU) as robots. A ROS-based layered software architecture, as shown in figure 2, drives MARSS hardware system. The Planner Layer has a planning executive and the AI planner \u2013 implemented through PlanSys2 \u2013 which connect and synchronize all sensor information, generate a planning problem on the fly based on pre-coded PDDL domain files, call the AI planner with the problem instance, and dispatch the plan. As it generates a plan, the AI planner is aware of the executable tasks available in the Task Layer. Tasks in the Task Layer are pre-coded through PDDL. A robot needs a set of basic skills to execute any task e.g., navigation, localization and those skills are the content of the Skill Layer. Skill Layer is implemented through integrating state-of-the-art robotics algorithms available through various ROS packages. The HIPAA-compliant security layer ensures that the robot system is resilient to network attacks (Privacy security) and robust to complete the required service (AI\/Software security). The privacy security module supports data encryption, access control, logging, backup, and firewall. A comprehensive penetration testing will be employed to evaluate and protect the safety of the network components. The software security module adopts fuzz testing, an advanced automated software testing technique that can rapidly search for states likely to cause system failures. The fuzzer runs for several rounds, where each round collects feedback from the previous round and aims at reaching deeper components that have not been tested before.<\/p>\n","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"class_list":["post-450","page","type-page","status-publish","hentry","post-no-thumbnail"],"_links":{"self":[{"href":"https:\/\/robotsforaging.cs.unh.edu\/index.php?rest_route=\/wp\/v2\/pages\/450","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/robotsforaging.cs.unh.edu\/index.php?rest_route=\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/robotsforaging.cs.unh.edu\/index.php?rest_route=\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/robotsforaging.cs.unh.edu\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/robotsforaging.cs.unh.edu\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=450"}],"version-history":[{"count":43,"href":"https:\/\/robotsforaging.cs.unh.edu\/index.php?rest_route=\/wp\/v2\/pages\/450\/revisions"}],"predecessor-version":[{"id":4322,"href":"https:\/\/robotsforaging.cs.unh.edu\/index.php?rest_route=\/wp\/v2\/pages\/450\/revisions\/4322"}],"wp:attachment":[{"href":"https:\/\/robotsforaging.cs.unh.edu\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=450"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}