Over the past few decades, DARPA’s Microsystems Technology Office (MTO) has enabled revolutionary advances in electronics materials, devices, and systems, which have provided the United States with unique defense and economic advantages. To continue its path of successful electronics innovation, DARPA has announced a new MTO effort called the Microsystems Exploration program. The Microsystems Exploration program will constitute a series of short-term investments into high-risk, high-reward research focused on technical domains relevant to MTO. Leveraging streamlined contracting and funding approaches, awards for each area of exploration – or μE topic – will be made within 90 days of announcement. Each μE topic will run for up to 18 months, during which time researchers will work to establish the feasibility of new concepts or technologies.
“This strategy of making smaller, targeted research investments will allow us to capitalize quickly on new opportunities and innovative research concepts,” said Dr. Mark Rosker, director of MTO. “The Microsystems Exploration program provides a way to assess whether or not a concept could evolve into a full program without requiring the use of more significant resources.”
The Microsystems Exploration program will employ best practices from DARPA’s other fast-track solicitation programs – the agency-wide AI Exploration program and the Defense Science Office’s “Disruptioneering” initiative. These programs are focused on enabling rapid advances in artificial intelligence and basic science respectively, and have shown numerous benefits to this approach. Similar to these efforts, the simplified proposal, contracting, and funding process employed by each μE topic will make it even easier for individuals and organizations to contribute to DARPA’s mission. Each award may be worth up to $1 million, as described in the individual μE solicitations.
To help advance MTO’s strategic imperatives, the Microsystems Exploration program will pursue innovative research concepts that explore frontiers in embedded microsystem intelligence and localized processing; novel electromagnetic components and technologies; microsystem integration for functional density and security; and disruptive microsystem applications in C4ISR, electronic warfare, and directed energy. In alignment with these technical domains, the first three potential topics focus on hardware security, novel materials, and new computing architectures for heterogeneous systems.
The first potential topic aims to address security issues within the hardware supply chain. Defense systems increasingly rely on commercial of the shelf (COTS) devices that move through complex supply chains, with each component changing hands several times. Throughout the process, nefarious actors have numerous opportunities to compromise the technology by introducing malicious circuitry – or hardware Trojans – to printed circuit boards (PCBs). The ability to detect when components are tampered with is difficult as the attacks are designed to remain hidden and avoid post-manufacturing tests until its functionality is triggered. The “Board-Level Hardware Security” related topic could explore the technological feasibility for real-time detection against these hardware Trojans installed in complex COTS circuit boards.
New uses of scandium (Sc)-doped aluminium nitride (AIN) could be investigated as a future potential μE topic. Sc-doped AlN is a popular material for a number of device applications, which span RF filters, ultrasonic sensors, and oscillators. Recent work has demonstrated the emergence of the material’s use in ferroelectric switching, which has enormous potential across a number of applications and devices. However, current exploration of this capability has been limited to a research setting. The “Ferroelectric Nitride Materials and Non-Volatile Memory” related topic could expand on this research, identifying the thickness and doping ranges that exhibit ferroelectric behavior, the robustness and reproducibility of the ferroelectric response, and further demonstrating ferroelectric nitrides as a technologically useful material.
Another potential μE topic could seek to address the trade-off between programmer productivity and performance that happens as hardware complexity continues to skyrocket. Advances at the hardware and software level that have enabled continued progress in computing performance, cost, and ubiquity have hit a wall. The expectation is that subsequent performance gains will come from an increased level of parallelism, specialization, and system heterogeneity, which will place further strain on programmer productivity. This “Massively Parallel Heterogeneous Computing” related topic could explore the creation of compiler technology that improves programmer productivity of massively parallel and heterogeneous processing systems.