Projects

Pamela

PAMELA is a probabilistic modeling language that extends the current state of the art process modeling languages, by adding first class probability variables. The language supports generative planning and temporal planning as well as supporting various learning algorithms for a more model-based approach than is possible with PDDL. Pamela is in use in numerous planning, learning, and social intelligence projects. Pamela can be compiled to produce HTN and TPN models and integrates with DOLL’s temporal and generative planners.

Learn more →

PAC2MAN

PAC2MAN is a project that allows distributed applications with network data flows to react to degraded network performance, either due to a concerted cyber-attack or for any other reason where network performance is compromised. A mission plan that encodes the applications « mission » in encoded in Pamela and a temporal planner tracks the progress of that mission while making changes that will side-step network issues in order to achieve the mission objectives.

Learn more →

DCRYPPS

DCRYPPS is an ongoing project to mobilize knowledge of systems designed using industry modeling languages, such as AADL and SYSML, to automatically find cyber vulnerabilities and to produce requirements for securing the design.

Learn more →

ADAM

On going project -- to be updated

Learn more →

RITA

On going project -- to be updated

Learn more →

KMASS

LISA is the Lightweight - A knowledge management system. The DOLL Team (DOLL and MIT CSAIL) is developing LISA, a knowledge management system that uses: Large Language Models (LLMs) to provide lightweight access to organizational knowledge.

Learn more →

Press Release

DOLL's News

DOLL Welcomes Thao Phuong to DOLL as Software Engineer

DOLL — 2020

place holder

DOLL/Vencore wins CHASE

DOLL — 2018

DOLL and MIT will subcontract to Vencore Labs on WILEE, a set of components for Threat Detection and Characterization as part of DARPA I2O's CHASE program. Scaling the Cyber Hunt problem poses several key technical challenges. Detecting and characterizing cyber threats at scale in large enterprise networks is hard because of the vast quantity and complexity of the data that must be analyzed as adversaries deploy varied and evolving tactics to accomplish their goals. WILEE's Threat Detector accelerates the hunting process by translating high-level threat descriptions into many possible concrete implementations and using adversarial planning, genetic perturbation, and data-driven evaluation to automatically prioritize hunt activities, inform data collection, minimize detection time, and detect both known and novel malicious activities with high confidence. WILEE's threat interpretation and validation process lower the cognitive burden on Cyber Protection Team (CPT) hunt operators by producing validated threat indicators in human-readable form, alongside a continuous confidence score to assist in triaging and prioritizing responses when operating with limited resources.

place holder

DOLL/Vanderbilt University wins CASE

DOLL — 2018

DOLL will subcontract to Vanderbilt University to develop DCRYPPS on the DARPA CASE program. DCRYPPS (Deriving Cyber-security Requirements Yielding Protected Physical Systems) will help automate the design and implementation of cyber-physical systems (CPS) to meet stringent cybersecurity requirements, without sacrificing functional performance, maintaining timing and cost constraints. The system will help derive the cybersecurity requirements, using model-based diagnosis, cyber-physical properties, and cyber-physical attack models, analysis of cyber requirements, and use of probabilistic certificates in managing cybersecurity requirements.

place holder

DOLL/SIFT wins CPS

DOLL — 2018

DOLL will subcontract to SIFT on the Creative Problem Solver (CPS) program, designing a system that will be able to generate and validate out-of-the-box approaches to a variety of difficult planning problems.  DOLL will begin its work on the program by designing a detailed simulation of some of the challenging problems described in Andy Weir's book, The Martian.

place holder

DOLL Team Demonstrates CART Robot-Assisted Repair Technology

DOLL — 2018

Developed under DARPA funding, DOLL has developed the capability for a robot assistant to guide a user in performing repair tasks using deep learning, active perception, mission representation, and plan recognition.  This video shows a user performing a simple repair operation consisting of replacing the battery of a cell phone, assisted by the CART system.

Youtube Link

place holder

DOLL wins phase I SBIR for hardware supported fast path planning for robots

DOLL — 2017

Under this contract, we will develop RAPID (Roadmap-based Accelerated Planning with Integrated Dynamics). The system will produce motion plans quickly, and it will also produce motion plans that permit a fast, agile motion for dynamic robot mechanisms.

place holder

DOLL wins Phase II STTR for its CART project with MIT

DOLL — 2016

In Phase II, DOLL will develop an integrated system that, based on a PAMELA-based mission plan for a detailed repair plan, will continuously observe the actions of the human performing the repair task.  DOLL will develop a machine-learning-based object detector that will dynamically determine the state and position of each of the items involved in the repair task, along with the state of the human's hands.  We will develop a belief state estimator that will make sense of these observations and determine their context and progress within the mission plan for the repair. In order to provide mixed-initiative guidance to the user, we will develop a voice interaction capability that will inform the user of the completion of each major step in the repair, inform the user of any mistakes, and allow the user to ask questions about their progress.

 

place holder

DOLL wins DARPA Distributed Autonomous Network Communication Enabler (DANCE) contract to develop autonomous quad-copter network

DOLL — 2016

DANCE will feature a self-organizing team of quadcopters (or a combination of quadcopters, unmanned winged aircraft, and ground robots) that implement a dynamic, ad hoc communication network that is robust to attacks and disturbances. These unmanned vehicles will be controlled by a mesh of automated distributed planners, guided by mission models.

place holder

DOLL was awarded a DARPA seedling to design and develop tools for military mission modeling

DOLL — 2016

This will leverage the PAMELA model representation to generate detailed Hierarchical Task Networks (HTNs) and Temporal Plan Networks (TPNs) that capture the Commander’s Intent along with all of the actionable details of the mission plan.  In addition, we will compare and reconcile classical computer science-focused planning with military planning.

place holder

DOLL team presents on PAMELA and PLANVIZ at Clojure/Conj

DOLL — 2016

DOLL team presents a session and a demonstration of PAMELA and PLANVIZ  for machine learning at Clojure/Conj 2016.

Youtube Link

place holder

DOLL team presents on PAMELA and PLANVIZ at Clojure West

DOLL — 2016

DOLL team presents a session and an unsession on PAMELA and PLANVIZ at Clojure/West.

DOLL/Vanderbilt University PAC2MAN as part of the DARPA EDGE-CT program

DOLL — 2015

DOLL will collaborate with MIT on a DARPA funded STTR contract called CART

DOLL — 2015

place holder

DOLL Welcomes Dan Cerys to DOLL as Chief Architect

DOLL — 2014

Dan Cerys joins DOLL effective January 1st 2014 as a Chief Architect.

place holder

DOLL Welcomes Andreas Hofmann to DOLL as Chief Scientist

DOLL — 2014

Andreas Hofmann joins DOLL effective May 1st 2014 as a Chief Scientist.

DOLL received a DARPA Phase II SBIR contract for a project called PAMELA

DOLL — 2014

DOLL and SIFT were awarded a DARPA seedling to explore active perception (Goshawk)

DOLL — 2014

DOLL/Applied Communications Sciences/Kestrel Institute PLANX

DOLL — 2013

DOLL Chief Scientist Robert Laddaga leaves DOLL to work at Vanderbilt

DOLL — 2013

DOLL welcomes Grace Robertson to DOLL as director of administration

DOLL — 2012

DOLL moves into new officess located in Lexington Massachusetts

DOLL — 2012

DOLL Chief Scientist Robert Laddaga is Co-Organizer of the Adaptive Host and Network Security Workshop at SASO

DOLL — 2012

DOLL Chief Scientist Paul Robertson is Co-General Chair of SASO 2012

DOLL — 2012

DOLL will work with MIT on their IARPA funded "StoneSoup" contract

DOLL — 2011

DOLL will subcontract to SIFT on a recently funded DARPA MRC project called STRATUS

DOLL — 2011

DOLL will subcontract to Applied Communications Sciences on a recently funded DARPA MRC project called ACCORD

DOLL — 2011

DOLL Welcomes Prakash Manghwani to DOLL as senior architect

DOLL — 2011

DOLL wins a DARPA seedling contract under the CRASH program: MARS

DOLL — 2010

place holder

Franz Inc and DOLL Inc win DARPA SBIR

DOLL — 1998

Smart Components: A High Level Semantically Based User Interface Management Tool for Distributed Applications: DARPA 96-062 SBIR Award 1998.

DOLL Inc wins DARPA award for advanced user interface capability

DOLL — 1996