Machine Learning

Just Round: Quantized Observation Spaces Enable Memory Efficient Learning of Dynamic Locomotion

Deep reinforcement learning (DRL) is one of the most powerful tools for synthesizing complex robotic behaviors. But training DRL models is incredibly compute and memory intensive, requiring large training datasets and replay buffers to achieve performant results. This poses a challenge for the next generation of field robots that will need to learn on the edge to adapt to their environment. In this paper, we begin to address this issue through observation space quantization. We evaluate our approach using four simulated robot locomotion tasks and two state-of-the-art DRL algorithms, the on-policy Proximal Policy Optimization (PPO) and off-policy Soft Actor-Critic (SAC) and find that observation space quantization reduces overall memory costs by as much as 4.2x without impacting learning performance.

Closing the Sim-to-Real Gap for Ultra-Low-Cost, Resource-Constrained, Quadruped Robot Platforms

As a step toward robust learning pipelines for these constrained robot platforms, we demonstrate how existing state-of-the-art imitation learning pipelines can be modified and augmented to support low-cost, limited hardware. By reducing our model’s observational space, leveraging TinyML to quantize our model, and adjusting the model outputs through post-processing, we are able to learn and deploy successful walking gaits on an 8-DoF, $299 (USD) toy quadruped robot that has reduced actuation and sensor feedback, as well as limited computing resources.

Tiny Robot Learning: Challenges and Directions for Machine Learning in Resource-Constrained Robots

Tiny robot learning lies at the intersection of embedded systems, robotics, and ML, compounding the challenges of these domains. This paper gives a brief survey of the tiny robot learning space, elaborates on key challenges, and proposes promising opportunities for future work in ML system design.

Machine Learning Sensors

Machine learning sensors represent a paradigm shift for the future of embedded machine learning applications. Current instantiations of embedded machine learning (ML) suffer from complex integration, lack of modularity, and privacy and security concerns from data movement. This article proposes a more data-centric paradigm for embedding sensor intelligence on edge devices to combat these challenges. Our vision for 'sensor 2.0' entails segregating sensor input data and ML processing from the wider system at the hardware level and providing a thin interface that mimics traditional sensors in functionality. This separation leads to a modular and easy-to-use ML sensor device. We discuss challenges presented by the standard approach of building ML processing into the software stack of the controlling microprocessor on an embedded system and how the modularity of ML sensors alleviates these problems. ML sensors increase privacy and accuracy while making it easier for system builders to integrate ML into their products as a simple component. We provide examples of prospective ML sensors and an illustrative datasheet as a demonstration and hope that this will build a dialogue to progress us towards sensor 2.0.

TinyML: Applied AI for Development

We believe that TinyML has a significant role to play in achieving the SDGs and facilitating scientific research in areas such as environmental monitoring, physics of complex systems and energy management. To broaden access and participation and increase the impact of this new technology, we present an initiative that is creating and supporting a global network of academic institutions working on TinyML in developing countries. We suggest the development of additional open educational resources, South–South academic collaboration and pilot projects of at-scale TinyML solutions aimed at addressing the SDGs.

COMS-BC3997-F22: Introduction to Robotics Engineering from Bits to Electrons

Robots are cyber-physical systems – leveraging computational intelligence to sense and interact with the real world. As such, robotics is a very diverse, cross-disciplinary field. This introductory course exposes learners to the vast opportunities and challenges posed by the interdisciplinary nature of robotics. While grounded and focused in computation this course also explores hands-on electromechanical and ethical topics that are an integral part of a real-world robotic system. Topics will include: a survey of the algorithmic robotics pipeline (perception, mapping, localization, planning, control, and learning), an introduction to cyber-physical system design, and responsible AI. The course will culminate in a team-based final project.

Mind the Gap: Opportunities and Challenges in the Transition Between Research and Industry

Mind the Gap: Opportunities and Challenges in the Transition Between Research and Industry is aimed at bridging the gap between academia and industry. For researchers, this workshop will help lift the curtain on the realities of academic to industry tech transfer. For industry experts, this workshop provides an opportunity to influence the direction of academic research. For both, we hope to provide an venue for integrated dialogue and identification of new potential collaborations.

TinyMLedu: The Tiny Machine Learning Open Education Initiative

[TinyMLedu](https://tinymledu.org) is working to build an international coalition of researchers and practitioners advancing TinyML in the developing world, and to develop and share high-quality, open-access educational materials globally.

Widening Access to Applied Machine Learning with TinyML

In this paper, we describe our pedagogical approach to increasing access to applied ML through a four part massive open online course (MOOC) on Tiny Machine Learning (TinyML) produced in collaboration between academia (Harvard University) and industry (Google). We suggest that TinyML, ML on resource-constrained embedded devices, is an attractive means to widen access because TinyML both leverages low-cost and globally accessible hardware, and encourages the development of complete, self-contained applications, from data collection to deployment. We also released the course materials publicly, hoping they will inspire the next generation of ML practitioners and educators and further broaden access to cutting-edge ML technologies.

HarvardX: Tiny Machine Learning MOOC

In this exciting Professional Certificate program offered by Harvard University and Google TensorFlow, you will learn about the emerging field of Tiny Machine Learning (TinyML), its real-world applications, and the future possibilities of this transformative technology. TinyML is a cutting-edge field that brings the transformative power of machine learning (ML) to the performance-constrained and power-constrained domain of embedded systems. The program will emphasize hands-on experience and is a collaboration between expert faculty at Harvard’s John A. Paulson School of Engineering and Applied Sciences (SEAS) and innovative members of Google’s TensorFlow team.

Harvard CS249r: Tiny Machine Learning (TinyML)

An introductory course on Applied AI at the intersection of Machine Learning and Embedded IoT Devices. We provide background on both topics and then dive into the unique challenges faced at that intersection point with hands-on assignments using TensorFlow, Google Colab, and Arduino.

Harvard CS249r: Special Topics in Edge Computing - Autonomous Machines

Modern embedded systems are intelligent devices that involve complex hardware and software to perform a multitude of cognitive functions collaboratively. Designing such systems requires us to have deep understanding of the target application domains, as well as an appreciation for the coupling between the hardware and the software subsystems.This course is structured around building “systems” for Autonomous Machines (cars, drones, ground robots, manipulators, etc.). For example, we will discuss what are all the hardware and software components that are involved in developing the intelligence required for an autonomous car?

Harvard CS 182: Introduction to Artificial Intelligence

Artificial Intelligence (AI) is an exciting field that has enabled a wide range of cutting-edge tech-nology, from driverless cars to grandmaster-beating Go programs. The goal of this course is to introduce the ideas and techniques underlying the design of intelligent computer systems. Topics covered in this course are broadly be divided into 1) planning and search algorithms, 2) probabilistic reasoning and representations, and 3) machine learning (although, as you will see, it is impossible to separate these ideas so neatly).