Convergence Accelerator Workshop - Chemical sensing with an olfaction analogue: high-dimensional, bio-inspired sensing and computation

生物嗅觉系统是地球上最精致的化学探测器。然而，与其他感官系统(如视觉、听觉)不同的是，它们的人工类似物给社会带来了革命性的变化(如图像和语音识别)，模仿嗅觉仍然是一个巨大的挑战。这在一定程度上是因为，为了识别气味，生物鼻子使用了大量的传感器--人类估计有400个传感器，每个传感器检测气味分子的不同性质。到目前为止，建造一个拥有如此多数量和多样化传感器的气味检测仪器是不可能的。但这种情况正在改变。在过去的十年里，出现了新的传感技术，使开发大型和多样化的化学传感器阵列成为可能。前景看好的技术包括那些使用与生物鼻子相同的传感器的技术，微型化技术，如纳米光子和微型机电系统，以及可以检测气味分子多种性质的传感器。与此同时，在模拟生物嗅觉系统中气味处理的算法以及高效处理大量数据的神经形态硬件方面也取得了重大进展。总之，这些进展为人工嗅觉和化学传感领域的新突破提供了巨大的潜力，直到最近，这些突破在技术上还是不可信的。然而，这需要将目前基本上独立于各学科(例如，生物化学、感知、生物工程、机器学习、医疗保健、制造)的不同专家小组聚集在一起。为了解决这一问题，一个跨学科的研究团队将组织一次互动研讨会，以(1)讨论即将转化为产品和应用的新兴传感和计算技术带来的机遇，以及(2)确定它们可以产生最大社会影响的应用领域，如环境监测、医疗保健、食品和饮料质量控制、精准农业和国土安全。这次研讨会的目标是为可能导致新产品和应用的人工化学感觉系统的翻译研究创建一个全面的3年路线图。研究团队将与一家辅导者公司合作，设计一个为期4天的研讨会，以小组和个人的形式进行各种汇聚和不同的活动。通过这些活动，将促使参与者确定人工化学传感系统的规格，确定最有希望的可供翻译的传感技术，以及可供应用和最终用户社区参与的技术。在整个研讨会中，参与者将被分成小组练习的小分组。这些小组可以是随机组成的，也可以根据一些重要因素(例如，专业知识的多样性、职业阶段、行业)进行设计。根据活动的不同，根据共性与多样性将参与者安排在分组讨论中可能是有益的。研究小组和协调小组将共同努力，为每项活动确定组织参与者的最佳方式。研讨会将被组织成一系列步骤，以“定义”需求，“确定解决方案”，然后“针对应用程序”。在第一步中，学员将为未来的化学传感器和传感器系统制定关键要求和规格。在步骤2中，与会者将确定五个领域的潜在方法：样品交付、生物传感器、微型分析仪器、多参数传感器和计算分析。最后，在国家3中，与会者将确定这项技术在2年、5年和10年可能产生社会影响的应用领域，并确定终端用户社区中哪些人应从事短期、中期和长期工作。参与者将从四个主要利益相关者中招募：工业、非营利组织、政府组织和学术界，涵盖范围广泛的学科，包括基础科学(材料科学、生物化学、感官感知、神经科学、分析化学)、工程(生物和神经形态工程、仪器、机器学习)、应用(医疗保健、环境、质量控制、农业、军事、安全)以及产品设计和制造。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

Biological olfactory systems are the most exquisite chemical detectors on the planet. Yet, unlike other sensory systems (e.g., vision, hearing), whose artificial analogues have revolutionized society (e.g., image and speech recognition), mimicking olfaction remains a grand challenge. This is in part due to the fact that, to recognize odors, biological noses use a large number of sensors –an estimated 400 sensors in humans, each sensor detecting different properties of odor molecules. Building an odor-detection instrument with such high number and diversity of sensors has not been possible to date. But this is changing. New sensing technologies have emerged over the past decade that make it possible to develop large and diverse arrays of chemical sensors. Promising technologies include those that use the same sensors as biological noses, miniaturization techniques such as nano-photonic and micro-electromechanical systems, and sensors that can detect multiple properties of an odor molecule. In parallel, there have been significant advances in algorithms to mimic odor processing in the biological olfactory system, and in neuro-morphic hardware to efficiently process large amounts of data. In concert, these advances offer tremendous potential for new breakthroughs in artificial olfaction and chemical sensing that until recently were technically implausible. However, this requires bringing together a diverse group of experts who currently work largely independently across disciplines (e.g., biochemistry, perception, bioengineering, machine learning, healthcare, manufacturing). To address this issue, a cross-disciplinary team of researchers will organize an interactive workshop to (1) discuss opportunities from emerging sensing and computing technologies that are ready to be turned into products and applications, and (2) identify application areas in which they can have the greatest societal impact, such as environmental monitoring, healthcare, quality control of food and beverages, precision agriculture, and homeland security. The goal of this workshop is to create a comprehensive 3-year roadmap for translational research in artificial chemo-sensory systems that may lead to new products and applications. The research team will collaborate with a facilitator company to design a 4-day workshop with a variety of convergent and divergent activities, to be performed in small groups and individually. Through these activities, participants will be prompted to define specifications for artificial chemosensory systems, identify the most promising sensing technologies that are ready for translation, and applications and end-user communities to engage. Throughout the workshop, participants will be divided into small breakouts for group exercises. These groups may be formed randomly or engineered based on some factor of importance (e.g., diversity of expertise, career stage, sector). Depending on the activity, it may be beneficial to place participants in breakout based on commonality vs diversity. The research team and the facilitation team will work together to determine, for each activity, the best way of organizing participants. The workshop will be structured into a series of steps to “define” the need, “identify the solution” and then “target applications.” In Step 1, participants will develop key requirements and specifications for the chemical sensors and sensor systems of the future. In Step 2, participants will identify potential approaches in five areas: sample delivery, biological sensors, micro-analytical instruments, multiparameter sensors, and computational analysis. Finally, in State 3, participants will identify application areas where this technology could have societal impact in 2-, 5- and 10-year timeframes, as well as identify those in the end-user communities who should be engaged in the short, medium and long term. Participants will be recruited from four major stakeholders: industry, non-profit organizations, government organizations, and academia, to cover a broad range of disciplines, including basic science (materials science, biochemistry, sensory perception, neuroscience, analytical chemistry), engineering (bio- and neuromorphic engineering, instrumentation, machine learning), applications (healthcare, environment, quality control, agriculture, military, security), as well as product design and manufacturing.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.