新方法让 AI 适用于安全关键场景

The “HardFlow” algorithm could help generative AI models produce high-quality outputs that obey strict requirements when “pretty close” doesn’t cut it. Adam Zewe | MIT News Publication Date: September 14, 2026 Press Inquiries Press Contact: Abby Abazorius Email: abbya@mit.edu Phone: 617-253-2709 MIT News Office Media Download ↓ Download Image Caption: HardFlow helps pretrained generative AI models satisfy hard constraints while improving solution quality without retraining, in applications spanning robotics, control of physical systems, and computer vision. Credits: Credit: MIT News; iStock *Terms of Use: Images for download on the MIT News office website are made available to non-commercial entities, press and the general public under a Creative Commons Attribution Non-Commercial No Derivatives license. You may not alter the images provided, other than to crop them to size. A credit line must be used when reproducing images; if one is not provided below, credit the images to "MIT." Close Caption: HardFlow helps pretrained generative AI models satisfy hard constraints while improving solution quality without retraining, in applications spanning robotics, control of physical systems, and computer vision. Credits: Credit: MIT News; iStock Previous image Next image MIT researchers have developed a new technique that helps generative artificial intelligence models find solutions to high-stakes problems. In these settings, a plausible answer is not enough: The output often must also satisfy nonnegotiable safety, physical, or task-specific requirements, known as hard constraints. The researchers developed a method that helps generative models meet these strict requirements without sacrificing the quality of their outputs. The key to their technique is to give the model more freedom during the generation process and enforce hard constraints on the final output, rather than at every intermediate step. In experiments spanning robotics, control of physical processes, and computer vision, the new method consistently satisfied the required constraints while identifying better solutions than existing techniques. This adaptable, plug-and-play technique works at deployment time, so it can be applied to pretrained generative models without retraining them. It can make such models more useful in applications where safety rules, physical laws, or other strict requirements cannot be violated. “The promise of generative AI is its ability to explore a rich space of possibilities, but the real world places boundaries on which possibilities are acceptable. Our approach lets us preserve that generative power while enforcing the nonnegotiable requirements of high-stakes or safety-critical applications,” says Navid Azizan, the Alfred H. and Jean M. Hayes Career Development Associate Professor in the Department of Mechanical Engineering and the Institute for Data, Systems, and Society (IDSS), a principal investigator of the Laboratory for Information and Decision Systems (LIDS), and the senior author of a paper on this technique. Azizan is joined on the paper by lead author Zeyang Li, a graduate student in mechanical engineering and LIDS; and Kaveh Alim, a graduate student in IDSS and LIDS. The research appears this week in the IEEE Transactions on Pattern Analysis and Machine Intelligence. Freedom to explore Pretrained generative AI models, such as diffusion models like Stable Diffusion and flow-matching models like FLUX, are now widely available. These powerful models learn to create new data by transforming random noise. Their availability has enabled people to adapt them to a wide range of applications. These highly capable models excel at providing answers that come close to satisfying most queries, but in safety-critical applications like robot path planning on a crowded factory floor, an answer that is “nearly correct” may not be good enough. For instance, a “nearly correct” path from one machine to another might still result in the robot colliding with a human co-worker. In such safety-critical applications, users often employ a technique called projection-based sampling, which repeatedly forces the model’s partial solutions, called intermediate samples, to satisfy strict requirements during the generation process. But constraining the entire generation process can prevent the model from reaching a better final solution. These methods also typically focus only on satisfying the hard constraints, missing the opportunity to improve other qualities of the solution, like reducing the length of the robot’s trajectory. “For constraint satisfaction, what ultimately matters is the model’s final output, since the internal process is discarded. By not requiring every intermediate step to satisfy the constraints, we give the model more freedom to find high-quality solutions that are still feasible in the end,” says Li. The researchers developed an algorithm called HardFlow that steers the sampling process so that the final output satisfies the user’s hard constraints without being overly restrictive and is of higher quality. Subtle steering HardFlow reformulates hard-constrained sampling as a trajectory-optimization problem, using tools from the field of optimal control. This enables the framework to steer the model’s sampling trajectory toward a goal, making subtle corrections along the way while enforcing hard constraints on the final output. “Control theory gives us a powerful framework for formalizing the optimal way of making these corrections,” Azizan says. But solving the trajectory-optimization problem around an enormous neural network was no easy task. The model may have hundreds of interconnected layers that process data. To make the problem tractable, the researchers leveraged the structure of flow-matching models to decompose the problem into a sequence of smaller, single-step subproblems. They then applied systematic transformations and approximations to derive an efficient, scalable algorithm that still finds a feasible solution. “Essentially, we transformed the trajectory-optimization problem into something that preserves the key properties of the original problem, but can be solved very efficiently at deployment time,” Azizan adds. Reformulating the task as an optimization problem allows HardFlow to incorporate additional goals that can improve the quality of the final answer. For instance, HardFlow could find a collision-free path for a robot that is also the shortest distance to its goal. “Our framework can jointly handle both aspects, which helps it perform much better than existing methods,” says Li. Across experiments in robotic manipulation, maze navigation, and text-guided image editing, HardFlow achieved perfect constraint satisfaction while consistently outperforming baseline methods on measures of solution quality. For example, it enabled a robotic manipulator to avoid collisions with obstacles while also finding the quickest path to the target object. Most other methods either resulted in collisions or found paths that took significantly more time. In addition, HardFlow’s computation time was comparable to or lower than that of most competing methods. In the future, the researchers could extend the framework to settings in which the AI model itself can also be updated, so that constraint satisfaction and sample quality can be improved in a more adaptive manner. Share this news article on: X Facebook LinkedIn Reddit Print Paper: “HardFlow: Hard-Constrained Sampling for Flow-Matching Models via Trajectory Optimization” Check for open access version(s) of the research mentioned in this article. Related Links Navid Azizan Laboratory for Information and Decision Systems Department of Mechanical Engineering Institute for Data, Systems, and Society Center for Computational Engineering School of Engineering MIT Schwarzman College of Computing Related Topics Research Computer science and technology Artificial intelligence Machine learning Algorithms Mechanical engineering Robotics Electrical engineering and computer science (EECS) Laboratory for Information and Decision Systems (LIDS) IDSS School of Engineering MIT Schwarzman College of Computing Related Articles A smarter way for large language models to think about hard problems AI-enabled control system helps autonomous drones stay on target in uncertain environments How to assess a general-purpose AI model’s reliability before it’s deployed A simpler method for learning to control a robot Previous item Next item MIT spinout turns plastic waste into resilient building materials Atlas Building Composites is commercializing MIT research to turn plastic waste into parts for buildings and other infrastructure. Read full story → Lifesaving Lincoln Laboratory device wins 2026 Excellence in Technology Transfer Award The handheld catheterization device AI-GUIDE, created by Lincoln Laboratory and Massachusetts General Hospital, promises improved health outcomes for injured service members and civilians. Read full story → MIT Schwarzman College of Computing launches pilot to help educators teach AI across disciplines A weeklong summer workshop brought higher education faculty to campus to explore how AI and machine learning materials can be adapted for their classrooms. Read full story → Injectable nanodevices could provide effective treatment for drug-resistant glioblastoma MIT researchers show implanted nanoantennas can be activated wirelessly to kill brain cancer cells without damage to healthy tissue. Read full story → A burst of “pink noise” may lead to more restorative sleep Delivered at just the right time, this type of auditory stimulus can strengthen the flow of cerebrospinal fluid, which clears debris from the brain and keeps it healthy. Read full story → An electrochemical approach turns ammonia into pure hydrogen An MIT team has demonstrated a more efficient way to extract pure hydrogen gas from hydrogen carrier molecules. Read full story →