爱丁堡大学将为大约 1/2的传感器和成像系统理学硕士学生提供理学硕士项目 。

格拉斯哥项目将通过格拉斯哥项目数据库进行选择。

下面列出了爱丁堡的项目。要申请分配给您的项目,请 通过电子邮件联系项目主管(请谷歌他们的名字以找到他们的电子邮件地址)。分配给您的项目后,请将相关信息(主管和项目名称)通过电子邮件发送给项目协调员 (johannes.courtial@glasgow.ac.uk)。

在爱丁堡大学举办的理学硕士 SIS 项目

Professor Tughrul Arslan (IMNS)

图格鲁尔·阿尔斯兰教授 (IMNS)

TA1 基于射频的导航 TA1 Radio Frequency based Navigation

全球导航卫星系统 (GNSS) 是最常见的户外定位模式。但是,它在某些地理环境中的性能仍然存在问题,并且容易受到故意干扰和欺骗以及无意干扰。一个在可用时补充 GNSS 并在不可用时替换的系统可能有助于填补这些空白。当 GPS 单元的高功耗是一个问题时,或者即使存在 GNSS 服务完全中断的风险,这种系统也可以独立于 GNSS 使用。因此,该项目将寻找替代方案,即使用已经存在于室外环境中的射频信号。所采取的方法将使用指纹识别。

The Global Navigation Satellite System (GNSS) is the most common mode of outdoor positioning. However, it still has problems with its performance in certain geographical environments and is prone to both intentional jamming and spoofing and unintentional interference. A system to supplement GNSS when available and replace when not, could be useful to fill these gaps. Such a system could also be used independent of GNSS for when the high-power consumption of a GPS unit is an issue, or even if there was a risk of total disruption of GNSS service. This project will therefore look into an alternative, using RF signals that are around in the outdoor environment already. The approach taken will be using fingerprinting.

用于机器人导航系统的 TA2 可重构计算

该项目的目标是在由微处理器和现场可编程门阵列 (FPGA) 组成的异构片上系统 (SoC) 板上开发机器人导航系统。FPGA 和硬核微处理器的配对允许重新配置计算,通过将 FPGA 的灵活性与硬核微处理器的处理能力相结合,实现更高效的导航系统。微处理器执行同时定位和映射 (SLAM) 算法,负责确定机器人的位置。SLAM 算法所需的数据首先由 FPGA 上实现的图像处理流水线进行处理。

TA3 用于微波头成像系统的柔性射频传感器的设计和制造

近年来,基于微波雷达的头部成像系统获得了大量曝光。在开发灵活且贴合人体头部的天线方面也取得了重大进展。石墨烯、液态金属、PDMS、硅树脂、导电纺织品和喷墨印刷技术等材料在该技术的许多应用中取得了进步。该项目的目标是为微波头成像系统设计和制造灵活的射频传感器。首先,将研究关于不同类型天线和不同类型材料的文献。然后,将在 CST studio 中设计和模拟符合要求的最佳天线设计。然后将制造天线并对其进行测试。

张丽贝卡教授 (IMNS)

RC1 微机电系统(3 个项目)

关键词:微电子,MEMS,设计,仿真

该项目涉及微机电系统 (MEMS) 的设计和仿真。MEMS 的主题很广泛,根据学生的兴趣,可以针对各种应用(例如能源、生物医学、环境等)设计和优化许多不同的 MEMS 设备。可以与学生讨论目标主题和规范。

Daniel Chitnis 博士 (IMNS)

DC1 使用 GPU 加速 SPICE 电路仿真

关键词:仿真、电路分析、编码、GPU

在任何电子电路和系统的开发中,一个关键的步骤是电路的模拟。类似于软件开发人员的调试器和作者的拼写检查器,仿真工具可以最大限度地减少电路设计人员的潜在错误,并为设计提供有效性。SPICE(集成电路重点仿真程序)于 1971 年在加州大学伯克利分校作为开源学习工具开始。今天,几乎所有开源和商业电路设计仿真软件都基于 1971 年的代码库,并进行了各种扩展和优化。现在是 SPICE 成立 50 周年,我们希望优化 SPICE,使其在 GPU 上的运行速度显着提高。CUDA 架构中提供的广泛并行性使加速大规模并行模拟和更快的灵敏度分析成为可能。

DC2 使用 FPGA 加速 SPICE 电路仿真

关键词:仿真、电路分析、编码、FPGA

开发电子电路和系统的关键步骤是电路的模拟。类似于用于软件开发的调试器和用于编写器的拼写检查器,仿真工具可以最大限度地减少设计人员的潜在错误并为设计提供有效性。SPICE(集成电路重点仿真程序)于 1971 年作为开源工具在加州大学伯克利分校开始学习。今天,几乎所有的开源和商业电路设计仿真都基于 1971 年的代码库,并进行了各种扩展和优化。现在是 SPICE 成立 50 周年,我们希望优化 SPICE 以显着提高在 FPGA 上的运行速度。FPGA 中提供的可重新配置硬件的实现灵活性,使加速某些 SPICE 内部计算成为可能。

使用 Xilinx PYNQ 平台的 DC3 强化学习

关键词:强化学习,FPGA,编码,PythonC++

人工智能最令人兴奋的应用之一是强化学习。这种类型的机器学习试图在没有先验知识的情况下解决复杂的现实世界问题。这意味着在学习过程中,算法需要完成大量的迭代以优化其模型参数。该项目研究在部署在边缘的 FPGA 上实现深度 Q 学习算法。我们使用 Xilinx PYNQ 平台在 Python 和 Vitis 中实现硬件加速,并使用 Open AI 的 Gym 环境来测试我们的实现。

DC4:用于实时模拟单光子雪崩二极管的数值模型

该项目的目标是为各种单光子雪崩二极管 (SPAD) 及其在成像(包括激光雷达和医疗保健)中的应用开发精确的数值模型。然后需要有效地实现模型,以便具有模拟真实硬件的实时性能。在评估 SPAD 在实际应用中的性能时,这种模型很有用。该项目涉及SPAD 操作的数值研究、JavaScript 和C++ 中的高效编码以及OpenGL 等图形框架。 

DC5:雪崩二极管的蒙特卡罗分析

该项目的目标是使用漂移扩散方程的 3D 建模对单光子雪崩二极管 (SPAD) 进行数值模拟。每个电子从光生成到电线的路径都被建模,以便为二极管内的电子运动创建最准确的统计模型。这使得可以针对每个 CMOS 工艺对二极管进行微调。该项目涉及半导体物理的一些基础知识,使用 MATLAB、Python 和 C++ 进行数值建模,以及结果的图形表示。

DC6:光子检测是先进的节点CMOS技术和2D设备

该项目研究了先进技术节点(如 5nm 和 3nm)对光子检测的适用性。虽然这些节点是针对数字电路的,例如智能手机中的日常处理器,但电子元件的小尺寸可能具有超快光子检测的潜力。此外,该项目还研究了更高级的节点,例如 2D 器件和单片晶体管。本项目涉及半导体物理基础知识,以及使用MATLAB和Python进行数值分析。

菲利普·汉斯博士 (IMNS)

PH1 用于极端环境中压力传感的法布里-珀罗干涉仪的设计和建模

关键词:光子学;传感器;微电子;微机电系统;设计造型。

该项目将与我们的工业合作伙伴 Druck Ltd 合作,测试用于极端温度和压力环境(例如喷气发动机或油井)的压力传感器的概念验证设计。学生将使用数学工具和仿真包来模拟和优化由碳化硅制成的假设微机电 (MEMS) 法布里-珀罗干涉仪的行为。将研究材料选择、尺寸、询问波长、温度和压力等变量,以优化信号响应、范围、灵敏度和可重复性。除了该项目之外,成功的设计将用于帮助竞标研究资金,以支持该设备的未来微制造和测试。

PH2 无线询问可穿戴压力传感器,用于医疗和运动紧身衣

关键词:射频电子;射频识别(RFID);微电子;可穿戴电子产品;传感器;设计造型

 该项目与赫瑞瓦特大学的学术合作伙伴和工业合作者 Semefab 合作,并与 Speedo 泳装和 Pentland Group 运动服装品牌合作,涉及用于监测的无线询问灵活可穿戴压力传感器的设计和性能建模医疗和运动压缩服装。该项目将建立在最近一位博士生的工作基础上,他设计并制造了一种新型柔性压力传感器设备,该设备由一个无源 RLC 电路组成,该电路的电谐振频率取决于施加的压力。目前,该设备使用电感耦合进行无线询问,但这种技术的检测范围非常有限,只有几毫米,并且容易出现未对准问题。项目学生将研究将射频识别 (RFID) 技术与传感器集成以在更远的距离内为设备供电并与设备进行通信并提高可靠性的可行性。该项目将主要是理论性的,涉及组合射频电路和微制造传感器系统的设计和性能建模。如果实验室准入,及时完成理论设计和测试的优秀学生也可以协助他们设计的制造和测试。关于组合射频电路和微制造传感器系统的设计和性能建模。如果实验室准入,及时完成理论设计和测试的优秀学生也可以协助他们设计的制造和测试。关于组合射频电路和微制造传感器系统的设计和性能建模。如果实验室准入,及时完成理论设计和测试的优秀学生也可以协助他们设计的制造和测试。

PH3激光竖琴

 关键词:嵌入式电子;微控制器;模拟和数字电子;光子学;软件图形用户界面;公众参与。

该项目旨在设计和建造一个交互式公众参与展览,以吸引年轻学生学习电子和电气工程。在未来的科学节和大学开放日展示,特定的展品将是激光竖琴。用户的手可以阻挡多束激光束,以产生不同的音符或触发音乐效果。距离传感器还支持额外的声音控制(音量、音高或其他效果)。其他传感器/执行器留给学生自己的想象力和判断力!Android 应用程序支持无线用户配置系统、录音、播放和其他有用的功能。最终的设备应该引人注目、使用简单且有趣、富有创意且具有智力吸引力。

PH4 视觉暂留 (POV) 显示

关键词:嵌入式电子;微控制器;数码电子;光子学;软件图形用户界面;公众参与。

该项目旨在设计和建造一个交互式公众参与展览,以吸引年轻学生学习电子和电气工程。具体的展览将是视觉暂留 (POV) 显示器。LED 的线性阵列围绕中心轴快速旋转,以绘制出 2 维或 3 维的可配置光体积。通过将旋转速度与 LED 寻址仔细同步,可以显示具有漂浮在自由空间中的错觉的 2D 或 3D 图像和动画。显示器可以是圆盘、圆柱体,甚至是球体。显示的进一步增强留给学生的想象力和判断力,并且最好包括用户交互性,例如简单游戏的集成,或自定义文本/图像的显示。

罗伯特·亨德森教授 (IMNS)

组织中脉冲光穿透的 RH1 建模

关键词:混合信号电子学,信号处理,光学,生物医学工程

时域近红外光谱是“光学 X 射线”的一种形式,通过这种形式,光以脉冲方式进入人体,并使用快速、灵敏的单光子探测器来检测光子穿过组织时的往返行程并为其计时。返回光子的时间历程揭示了有关体内深处血流的信息。这是其他类型的身体扫描(如 X 射线或正电子发射断层扫描)的一种有前途的低成本、无辐射替代方案。在这个项目中,学生将使用 Matlab 对光子时间历程进行建模,并研究用于捕获该信号的不同时间到数字转换器 (TDC) 的安排。该项目与 Quantic 项目(量子增强成像中的 Quantum Hub)有关,目前正在为此设计一种新的 CMOS 检测器芯片。

RH2 自适应时间门控线扫描激光雷达

关键词:数字电子、Verilog FPGA 设计、光学、光子学、CMOS SPAD 传感器

由于对自动驾驶汽车的兴趣和工业自动化水平的提高,光探测和测距 (LIDAR) 正在享受一段巨大的投资时期。该项目将采用在爱丁堡设计的具有 512 个定时通道的新型单光子雪崩二极管 (SPAD) 线传感器,以在 100 米范围内执行激光雷达。学生将从事 Verilog 代码的开发,以激活和试用传感器中新的自适应门控设施。这允许每个像素列中的时间门了解目标的距离并抑制背景光,否则会破坏信号。根据项目时的限制,工作可能完全基于编码或可能涉及一些实验室 LIDAR 测量。

RH3 InSb 雪崩光电二极管建模和电容跨阻放大器设计

关键词:模拟电子,Cadence Design Systems,激光雷达,雪崩光电二极管

由于对自动驾驶汽车的兴趣和工业自动化水平的提高,光探测和测距 (LIDAR) 正在享受一段巨大的投资时期。一个有希望的方向是将这些系统移向短波红外 (SWIR) 中的检测波长,其中正在开发 InSb 或 Si on Ge 检测器。该项目将开发谢菲尔德大学正在开发的新型 InSb 雪崩光电二极管的 Cadence Spectre VerilogA 宏模型。这些模型将用于制作具有单光子灵敏度的电容跨阻放大器原型。学生将对合适的电路结构进行调查,并使用 Cadence Spectre 对其进行评估。与为 SPAD 设备开发的现有电子设备的兼容性或可重构性将是有利的。

Srinjoy Mitra 博士 (IMNS)

SM1 关键工人无线温度传感器

关键词:固件编程,温度传感器,PCB设计,RFID,无线通信

核心体温升高 (CBT) 是 Covid-19 的一个关键指标:它是最早和最常见的症状,在任何其他症状出现后几天最常出现发烧。使用传统温度计进行间歇性温度监测会因密切接触而存在感染风险,并且对临床和护理部门而言都是劳动密集型的。红外枪和热像仪对于繁忙地区的工作人员(尤其是佩戴 PPE)有太多限制,并且不适用于视线和技术培训可能具有挑战性的护理部门。我们计划设计一种基于 RFID 的温度传感器,该传感器将放置在内耳中,用户可以全天佩戴。当用户距离位于建筑物内多个位置的阅读器 2m 以内时,将读取温度读数。

用于生物阻抗监测和电刺激的 SM2 可穿戴设备(2 个项目)

关键词或主题领域:生物电势、生物阻抗、PCB 设计、

生物阻抗监测可以提供许多临床相关的信号,这些信号可能对疾病的早期诊断有用。一种称为聚焦阻抗的新方法可以检测特定器官(例如肺、膀胱)的变化,并可用于新型可穿戴传感器。我们将使用商用阻抗监测芯片来设计一个便于用户操作的便携式生物阻抗系统。生物阻抗测量通常需要四个电极,两个用于电流源,两个用于测量合成电压。一些生物阻抗信号很容易在您的身体上测量(例如,心跳期间的变化)并将用于演示系统。我们还将尝试使用电流源来创建便携式电刺激功能性电刺激 (FES),这是治疗某些健康状况的非常有用的工具。

斯图尔特·史密斯博士 (IBioE)

用于肺部健康的 SS1 传感器

关键词:生物医学、传感、微控制器

在这个项目中,您将与医学院和兽医学院的同事合作开发一个新系统来测量慢性阻塞性肺病 (COPD) 和其他长期呼吸系统疾病患者的肺功能。这是 Proteus 项目外展活动的一部分,鼓励 COPD 患者加入歌唱团体(The Cheyne Gang),学习如何控制呼吸和改善肺功能。我们希望您开发一种新的传感器系统来测量肺功能,它基于您可能熟悉的“峰值流量计”之类的东西。理想情况下,这将进行可靠的测量,这些测量可以以电子方式存储或发送到移动设备以监控和跟踪变化。新设备将与临床医生和患者合作开发。

博物馆中的 SS2 灯

关键词:混合信号电子学,可见光通信,公众参与

我们希望与通信博物馆 (http://mocft.co.uk) 合作开发一个可见光通信系统示例,用于向参观者提供有关博物馆展品的信息。这个概念是在展览上使用 LED 照明将数据传送到参观者携带的手持设备。该项目将涉及开发传输系统、调制 LED 照明、手持设备和交付方法。示例可能包括文本和图像的传递、音频评论,并且媒体选择将决定整个系统。成功的系统可以作为特别展览的一部分在博物馆进行试验。

用于集成生物传感器或半导体测量自动化的 SS3 电化学测试台

关键词:软件开发、硬件控制、生物传感器、半导体

该项目将涉及开发软件来控制半导体晶圆探测器,以在晶圆级测量传感器或半导体器件。对于电化学希望自动化这些测量,将电气/电子测量与液体处理系统相结合,以分配电解质进行电化学测量。这也可能涉及自动化半导体器件和微型测试结构的更多标准电气测量,或微机电器件的光学测量。

亚当斯托克斯博士 (IMNS)

AS1 探索微电子与微流体的协同设计

关键词:微电子学、微制造、生物电子学、数字系统、模拟系统。 该项目将探索微电子和微流体协同设计的最新进展,学生将设计测试结构来探索,例如,片上冷却、片上流体布线和自上而下的设计流程——摩尔微电子系统。理想情况下,该项目将在 IMNS 洁净室中进行一些制造,但也可以作为模拟和设计任务完成。学术文献的一个很好的起点是最近的自然论文:https://www.nature.com/articles/d41586-020-02503-1

AS2 液体计算机

关键词:微电子学,微制造,生物电子学,数字系统,模拟系统

在这个项目中,我希望学生回顾有关液态计算机的文献。正如最近一篇评论论文的作者所指出的那样:“计算的基板不一定是固体。完全用液体制造计算机是可能的。” (https://royalsocietypublishing.org/doi/10.1098/rstb.2018.0372)这个项目可能是纯理论的,或者如果需要/可能的话,可以包含一个制造元素。理想的学生应该对数字系统设计有扎实的理解,并且愿意探索跨学科研究,包括流体力学、生物计算、分布式计算架构和复杂性科学等等。

乔纳森·特里博士 (IMNS)

JT1工艺与器件仿真辅助微电子教学(3个项目)

关键词:微电子器件,仿真,高等教育教学

项目学生将建立一个实验室,以协助向本科生教授微电子器件理论。Synopsys Sentaurus Workbench 软件平台将用于开发基于模拟的练习,以帮助 UG 学生了解微电子器件的理论和概念以及它们的制造对其性能的影响。以下为建议选题,但项目学生可自行选择适合UG课程的领域:

• MOSFET 器件中的短沟道效应

• 双极结型晶体管

• 微电子设备构建块

 主要目标将是开发完整的学习体验,以补充提供给 UG 队列的讲座和其他材料。

Istvan Gyongy 博士 (IMNS)

使用单光子计数传感器的 IG1 光学血压监测

关键词:信号处理,生物医学传感,可穿戴技术

近年来,用于测量心率或血氧等生理参数的可穿戴设备激增。然而,连续的、无袖带的血压监测仍然是一个挑战,特别是在紧凑、便携式仪器的形式中。最有前途的技术之一是测量脉冲传输时间 (PTT),即动脉压力波到达身体某个部位的时间。发现 PTT 与血压密切相关,但需要对个别受试者进行校准。

该项目旨在应用工程学院开发的单光子敏感 (SPAD) 图像传感器,通过光学方式测量 PTT。光纤将用于将图像传感器与 LED 照明源一起耦合到身体的两个不同位置。然后将设计适当的信号处理方案来分析由图像传感器捕获的反射信号,并提取 PTT。

用于单光子激光雷达的 IG2 MEMS 扫描

关键词:3D成像,MEMSFPGA

3D 飞行时间/激光雷达设备正在成为许多机器人和自主系统(例如自动驾驶汽车)的关键组件。然而,实现大视场 (FOV) 与高帧速率相结合可能很困难,这会影响此类系统在高度动态环境中的感知。

鉴于此问题,该项目将探索在使用单光子敏感 (SPAD) 图像传感器的扫描 3D 飞行时间/LIDAR 成像中使用先进的 MEMS 反射镜。工程学院开发的现有 SPAD 传感器将与现成的 MEMS 反射镜结合使用,从而实现点对点扫描。FPGA 将被编程以控制 SPAD 和 MEMS 设备。将考虑单基地(SPAD 与激光照明同轴)和双基地(SPAD 具有广角光学器件,仅扫描激光)配置以及智能(场景相关)扫描方案。

使用闪光灯、直接飞行时间传感器的 IG3 3D 成像

关键词:3D 成像,信号处理,神经网络

具有 3D 成像功能的智能手机和平板电脑通常使用基于连续波照明的间接飞行时间 (iToF) 传感器和具有光解调器像素的接收器。然而,苹果和三星最近的设备采用了直接飞行时间 (dToF) 阵列,其中每个像素都能够与脉冲光源结合检测单个光子并为其计时。这些阵列生成高精度的深度图,不受多路径反射的影响,这是 iToF 系统中的一个常见问题,导致某些场景的严重不准确。当前 dToF 成像器的主要缺点是阵列尺寸有限,因此横向 (xy) 分辨率有限。因此,该项目将考虑基于高分辨率强度数据放大 dToF 深度数据。

伊恩·安德伍德教授 (IMNS) https://www.eng.ed.ac.uk/about/people/prof-ian-underwood

IU1 3D影像相关项目(3个项目)

主题将涉及:

• 使用 3D 光学传感

o STMicro Flightsense 传感器。

英特尔实感摄像头

o 集成在智能手机中的传感器

• 使用 3D 数据可视化

o 头戴式显示器

o 光场显示

博士。Filippo Menolascina (IBioE)

用于酵母细胞分割的 FM1 深度学习

在本项目中,您将使用 U-Net,这是一种深度人工神经网络,可在微流体设备的显微图像中自动识别酵母细胞。该项目的目的是在 U-Net 的超参数空间中进行搜索,以最大限度地减少验证错误。本项目需要对矩阵代数和 Python 编程有基本的了解。

用于大肠杆菌细胞分割的 FM2 深度学习

在本项目中,您将使用 U-Net,这是一种深度人工神经网络,可自动识别微流体设备显微图像中的细菌细胞。该项目的目的是在 U-Net 的超参数空间中进行搜索,以最大限度地减少验证错误。本项目需要对矩阵代数和 Python 编程有基本的了解。  

杰米·马兰博士 (IMNS)

JM1 光学组织氧合监测

关键词:生物医学,光学传感,微控制器,转化研究

体内组织的氧合是必不可少的,但在医院手术后可能会由于局部血液供应故障而受到影响。这可能会给患者带来危险的并发症。该项目的目的是探索使用微型光学氧传感器(类似于可穿戴健身追踪器中的传感器)进行组织氧合的植入和术后监测。它将涉及编写用于传感器和微控制器之间通信的固件,以及开发用于向临床医生显示数据的可视化界面。将有机会与 NHS 执业外科医生共同设计显示系统。这项工作将为该技术在临床前体内 模型中的未来试验奠定基础 。

最后修改时间:2021 年 3 月 30 日,星期二,下午 12:19

MSc SIS Projects hosted at the University of Edinburgh

Professor Tughrul Arslan (IMNS)

TA1 Radio Frequency based Navigation

The Global Navigation Satellite System (GNSS) is the most common mode of outdoor positioning. However, it still has problems with its performance in certain geographical environments and is prone to both intentional jamming and spoofing and unintentional interference. A system to supplement GNSS when available and replace when not, could be useful to fill these gaps. Such a system could also be used independent of GNSS for when the high-power consumption of a GPS unit is an issue, or even if there was a risk of total disruption of GNSS service. This project will therefore look into an alternative, using RF signals that are around in the outdoor environment already. The approach taken will be using fingerprinting.

TA2 Reconfigurable Computing for Robotic Navigation Systems

The aim of this project is to develop a robotic navigation system on a heterogeneous System-on-Chip (SoC) board that consists both a microprocessor and Field-Programable Gate Array (FPGA). The pairing of the FPGA and a hard-core microprocessor allows for reconfigurable computing, achieving a much more efficient navigation system by combining the flexibility of FPGAs with the processing power of a hard-core microprocessor. The microprocessor executes Simultaneous Localisation and Mapping (SLAM) algorithms that are responsible for determining the location of the robot. The data required for the SLAM algorithms are first processed by the image processing pipeline implemented on the FPGA.

TA3 Design and fabrication of flexible RF sensors for microwave head imaging systems

Microwave radar-based imaging systems for the head have gained significant exposure in recent years. There have also been significant work in developing antennas that are flexible and conformable to the human head. Materials such as graphene, liquid metal, PDMS, silicone, conductive textile, and inkjet printing techniques have generated advancement in this technology for many applications. The goal of this project is to design and fabricate flexible RF sensors for microwave head imaging systems. Firstly, literature about different types of antenna and different types of materials will be studied. Then, the optimal antenna design that meets requirements will be designed and simulated in CST studio. The antennae will then be fabricated and have them tested.

Professor Rebecca Cheung (IMNS)

RC1 Microelectromechanical Systems (3 projects)

Keywords: Microelectronics, MEMS, Design, Simulation

The project is concerned with the design and simulation of microelectromechanical systems (MEMS). The topic of MEMS is broad and depending on the student’s interest, a number of different MEMS devices can be designed and optimised for various applications, for example, energy, biomedical, environmental and more. The targeted topic and specification can be discussed with the student.

Dr Danial Chitnis (IMNS) https://www.eng.ed.ac.uk/about/people/dr-danial-chitnis

DC1 Accelerating SPICE circuits simulation using GPU

Keywords: Simulation, Circuit Analysis, Coding, GPU

A crucial step in development of any electronic circuit and system, is the simulation of the circuits. Similar to a debugger for a software developer, and a spell checker for a writer, a simulation tool minimises the potential errors of the circuit designer and provides validity to the design. SPICE (Simulation Program with Integrated Circuit Emphasis) started as an opensource learning tool back in 1971 at UC Berkeley. Today, almost all the opensource and commercial circuit design simulation software are based on the 1971 codebase with various extensions and optimisations. Now on its 50th anniversary, we want to optimise SPICE to run significantly faster on a GPU. The extensive parallelism offered in CUDA architectures enables the possibility of accelerating massively parallel simulations and quicker sensitivity analysis. This project investigates which parts of the circuit analysis benefits from GPU’s parallelism by implementing the code using NVIDIA’s CUDA C++ and Python compilers.

DC2 Accelerating SPICE circuits simulation using FPGA

Keywords: Simulation, Circuit Analysis, Coding, FPGA

A crucial step in developing and electronic circuit and system, is the simulation of the circuits. Similar to a debugger for a software develop, and spell checker for writer, a simulation tool minimises the potential errors of the designer and provides validity to the design. SPICE (Simulation Program with Integrated Circuit Emphasis) started a learning back 1971 as an opensource tool at UC Berkeley. Today, almost all the opensource and commercial circuit design simulation are based on the 1971 codebase with various extenuations and optimisation. Now on its 50th anniversary, we want to optimise SPICE to run significantly faster on an FPGA. The flexibility of implementing re-configurable hardware provided in an FPGA, enables the possibility of accelerating some of SPICE internal computations. This project investigates which parts of the circuit analysis can be implemented inside an FPGA and benefit from the acceleration using the Xilinx’s new Vitis v++ compiler.

DC3 Reinforcement learning using Xilinx PYNQ platform

Keywords: Reinforcement learning, FPGA, Coding, Python, C++

One of the most exciting applications of artificial intelligence is reinforcement learning. This type of machine learning tries to solve complex real-world problems without a priori knowledge. This means during the learning process, the algorithm is required to complete a significant number of iterations to optimise its model parameters. This project investigates implementation of a deep Q-learning algorithm on an FPGA deployed at the edge. We use the Xilinx PYNQ platform for implementing the hardware acceleration in Python and Vitis, and Open AI’s Gym environment for testing our implementation.

DC4: Numerical model for real-time simulation of single photon avalanche diodes

The goal of this project is to develop an accurate numerical model for a variety of single photon avalanche diodes (SPAD) and their application in imaging including LIDAR and healthcare. The model then needs to be implemented efficiently in order to have a real-time performance emulating a real hardware. Such a model is useful when evaluating the performance of SPADs in real world applications. This project involves numerical study of the SPAD operation, efficient coding in JavaScript and C++, and graphical frameworks such as OpenGL. 

DC5: Monte-Carlo analysis of avalanche diodes

The goal of this project is performing a numerical simulation for single photon avalanche diodes (SPAD) using 3D modelling of drift-diffusion equations. The path of each electron from photo generation to electrical wiring is modelled in order to create the most accurate statistical model for movement of electrons within the diode. This enables fine tuning of the diode for each CMOS process. This project involves some basic knowledge of semiconductor physics, numerical modelling using MATLAB, Python and C++, and graphical representation of the results.

DC6: Photon detection is advanced node CMOS technology and 2D devices

This project investigates the suitability of the advanced technology nodes (such as 5nm and 3nm) for photon detection. Although these nodes are targeted for digital circuits such as the everyday processor in a smartphone, the small size of the electrical components may have the potential for ultrafast photon detection. In addition, this project investigates more advanced nodes such as 2D devices and single sheet transistors. This project involves a basic knowledge of semiconductor physics, and numerical analysis using MATLAB and Python.

Dr Philip Hands (IMNS)

PH1 Design and modelling of Fabry-Perot interferometers for pressure sensing in extreme environments

Keywords: Photonics; Sensors; Microelectronics; MEMS; Design; Modelling.

In collaboration with our industrial partner, Druck Ltd, this project will test a proof-of-concept design for a pressure sensor, for use in extreme temperature and pressure environments (such as jet engines or oil wells). The student will use mathematical tools and simulation packages to model and optimise the behaviour of a hypothetical microelectromechanical (MEMS) Fabry-Perot interferometer, made from silicon carbide. Variables such as choice of materials, dimensions, interrogation wavelength, temperature and pressure will be investigated, with a view to optimizing signal response, range, sensitivity and repeatability. Beyond the project, a successful design will be used to help bid for research funding to support the future microfabrication and testing of the device.

PH2 Wirelessly interrogated wearable pressure sensors for medical and sports compression garments

Keywords: RF electronics; Radio frequency identification (RFID); Microelectronics; Wearable electronics; Sensors; Design; Modelling

 In collaboration with academic partners from Heriot-Watt University and with industrial collaborators Semefab, and also in association with Speedo swimwear and the Pentland Group of sportswear brands, this project concerns the design and performance modelling of wirelessly interrogated flexible and wearable pressure sensors for the monitoring of medical and sports compression garments. The project will build upon the work of a recent PhD student, who designed and fabricated a novel flexible pressure sensor device consisting of a passive RLC circuit with an electrical resonant frequency that depends upon applied pressure. Currently the device is interrogated wirelessly using inductive coupling, but this technique has a very limited range of detection of only a few mm and is prone to misalignment issues. The project student will investigate the feasibility of integrating radio frequency identification (RFID) technology with the sensor to power and communicate with the device over greater distances and with improved reliability. The project will be mostly theoretical, concerning design and performance modelling of the combined RF circuit and microfabricated sensor system. If laboratory access permits, an exceptional student who completes theoretical design and testing in good time, may also be able to assist in the fabrication and testing of their design.

PH3 Laser harp

 Keywords: Embedded electronics; Microcontrollers; Analogue and digital electronics; Photonics; Software GUI; Public Engagement.

This project seeks to design and build an interactive public engagement exhibit for the purpose of attracting young students to study electronics and electrical engineering. Demonstrated at future science festivals and University open days, the specific exhibit will be that of a laser harp. Multiple laser beams can be blocked by the user’s hands to induce different notes or to trigger musical effects. Distance sensors also enable additional sound control (volume, pitch or other effect). Other sensors/ actuators are left to the student’s own imagination and discretion! An Android app enables wireless user configuration of the system, recording, playback and other useful features. The final device should be eye-catching, simple and fun to use, creative and intellectually engaging.

PH4 Persistence of vision (POV) display

Keywords: Embedded electronics; Microcontrollers; Digital electronics; Photonics; Software GUI; Public Engagement.

This project seeks to design and build an interactive public engagement exhibit for the purpose of attracting young students to study electronics and electrical engineering. The specific exhibit will be that of a persistence-of-vision (POV) display. A linear array of LEDs are rapidly rotated around a central axis to map out a 2 or 3-dimensional volume of configurable light. With careful synchronisation of the rotation speed with the LED addressing, 2D or 3D images and animations can be displayed that have the illusion of floating in free-space. The display could be a circular disk, a cylinder, or even a sphere. Further enhancement of the display is left to the student’s imagination and discretion, and would ideally include user interactivity, such as the integration of a simple game, or the display of custom text/images.

Prof Robert Henderson (IMNS)

RH1 Modelling of Pulsed Light Penetration in Tissue

Keywords: Mixed Signal Electronics, Signal Processing, Optics, Biomedical Engineering

Time-domain near infrared spectroscopy is a form of “Optical X-Ray” whereby light is pulsed into the body and fast, sensitive single photon detectors are used to detect and time the round trip of photons as they pass through tissue. The time history of the returning photons reveals information about blood flow deep inside the body. This is a promising low-cost, radiation-free alternative to other types of body scanning such as X-ray or Positron Emission Tomography. In this project the student will use Matlab to model the photon time history and study different time to digital converter (TDC) arrangements for capturing this signal. The project is related to the Quantic Project (Quantum Hub in Quantum Enhanced Imaging) where a new CMOS detector chip is currently being designed for this purpose.

RH2 Adaptive Time Gated Line Scanning LIDAR

Keywords: Digital Electronics, Verilog FPGA Design, Optics, Photonics, CMOS SPAD Sensors

Light Detection and Ranging (LIDAR) is enjoying a period of enormous investment fuelled by interest in self-driving vehicles and increasing levels of industrial automation. This project will employ a new single photon avalanche diode (SPAD) line sensor designed in Edinburgh with 512 timing channels to perform LIDAR in a few 100m range. The student will work on development of Verilog code to activate and trial a new adaptive gating facility in the sensor. This allows a time gate in each pixel column to learn the distance of the target and suppress background light which would otherwise corrupt the signal. Depending on restrictions at the time of the project the work may be entirely coding based or may involve some laboratory LIDAR measurements. The project is aligned to two industry-oriented research projects being conducted in the department (SPEXS and SPIDAR) and will be supported by related researchers.

RH3 InSb Avalanche Photodiode Modelling and Capacitive Transimpedance Amplifer  Design

Keywords: Analogue Electronics, Cadence Design Systems, LIDAR, Avalanche Photodiodes

Light Detection and Ranging (LIDAR) is enjoying a period of enormous investment fuelled by interest in self-driving vehicles and increasing levels of industrial automation. A promising direction is to move these systems towards detection wavelengths in the short wave infrared (SWIR) where InSb or Si on Ge detectors are being developed. The project will develop Cadence Spectre VerilogA macromodels of new InSb avalanche photodiodes being developed at Sheffield University. The models will be used to prototype capacitive transimpedance amplifiers with single photon sensitivity. The student will conduct a survey of suitable circuit structures and evaluate these using Cadence Spectre. Compatibility or reconfigurability with existing electronics being developed for SPAD devices will be advantageous. The circuits will be designed for a 2D LIDAR pixel matrix in a suitably advanced CMOS technology node

Dr Srinjoy Mitra (IMNS)

SM1 Wireless temperature sensor for key-workers

Keywords: Firmware programming, Temperature sensor, PCB design, RFID, wireless communication

Elevated core body temperature (CBT) is a key indicator of Covid-19: it is the earliest and most common symptom, with fever presenting most frequently days after onset of any other symptoms. Intermittent temperature monitoring using traditional thermometers have infection risk through close contact and is labour intensive for both clinical and care sectors. Infra-red guns and thermal cameras have too many limitations for staff in busy areas (particularly wearing PPE) and are poorly applicable to care sectors where line of sight and technical training may be challenging. We plan to design a RFID based temperature sensor that will be placed in the inner-ear and can be worn by the user throughout the day. The temperature reading will be taken when the user is within 2m of a reader positioned at multiple locations within the building.

SM2 Wearable device for bio-impedance monitoring and electrical stimulation (2 projects)

Keywords or subject area: Bio-potential, bio-impedance, PCB design,

Bio-impedance monitoring can provide a number of clinically relevant signals that could be useful for early diagnosis of disease. A new method called focused-impedance can detect changes in specific organs (e.g. lungs, bladder) and can be used for novel wearable sensors. We will use commercial impedance monitoring chips to design a portable bio-impedance system that can be easily handled by the user. Bio-impedance measurement normally requires four electrodes, two for current sources and two for measuring the resultant voltage. Some bio-impedance signals are very easy to measure on your body (e.g., changes during heartbeat) and will be used to demonstrate the system. We will also try to use current sources to create portable electrical stimulus functional-electrical-stimulation (FES), a very useful tool for treating certain health conditions.

Dr Stewart Smith (IBioE)

SS1 Sensors for Lung Health

Keywords: Biomedical, Sensing, Microcontrollers

In this project you will work with colleagues in the College of Medicine and Veterinary Medicine to develop a new system to measure lung function in patients with Chronic Obstructive Pulmonary Disease (COPD) and other long term respiratory conditions. This is part of the outreach activity from the Proteus project, where patients with COPD have been encouraged to join a singing group (The Cheyne Gang) to learn how to control their breathing and improve their lung function. We would like you to develop a new sensor system to measure lung function, based on something like the “peak flow meter” that you may be familiar with. Ideally this would make reliable measurements which can be stored electronically or sent to a mobile device for monitoring and tracking changes. The new device would be developed in collaboration with clinicians and patients.

SS2 Light in the Museum

Keywords: Mixed signal electronics, visible light communication, public engagement

In conjunction with the Museum of Communication (http://mocft.co.uk), we would like to develop an example of a visible light communication system for use in delivering information about exhibits in the museum to visitors. The concept is to use LED lighting over an exhibit to deliver data to a handheld device carried by the visitor. The project would involve developing the transmission system, to modulate the LED lighting, the handheld device, and the method of delivery. Examples could include delivery of text and images, audio commentary, and the media choice will determine the overall system. Successful systems could be trialed at the museum as part of a special exhibition.

SS3 Electrochemical test bench for integrated biosensors or automation of semiconductor measurements

Keywords: Software development, Hardware control, Biosensors, Semiconductors

This project would involve developing software to control a semiconductor wafer prober to make measurements of sensors or semiconductor devices at the wafer level. For electrochemical want to automate these measurements which combine electrical/electronic measurements with a liquid handling system to dispense electrolytes for electrochemical measurements. This could also involve automating more standard electrical measurements of semiconductor devices and microscale test structures, or optical measurements of microelectromechanical devices.

Dr Adam Stokes (IMNS)

AS1 Exploring the Co-Design of Microelectronics and Microfluidics

Keywords: microelectronics, microfabrication, bioelectronics, digital systems, analogue systems. This project will explore recent progress in the co-design of microelectronics and microfluidics and the student will design test structures to explore, for example, on-chip cooling, on-chip fluidic routing and the top-down design flow of more-than-moore microelectronic systems. Ideally this project would have some fabrication in the IMNS cleanroom, but it can also be completed as a simulation and design task. A good starting point in the academic literature is this recent nature paper: https://www.nature.com/ articles/d41586-020-02503-1

AS2 Liquid Computers

Keywords: microelectronics, microfabrication, bioelectronics, digital systems, analogue systems

In this project I would like the student to review the literature around liquid computers. As the authors of a recent review paper note “A substrate does not have to be solid to compute. It is possible to make a computer purely from a liquid.” (https://royalsocietypublishing.org/doi/10.1098/rstb.2018.0372) This project could be purely theoretical, or could contain a fabrication element if desired/possible. The ideal student would have a solid understanding of digital systems design, and also a willingness to explore interdisciplinary research including fluid mechanics, biological computation, distributed computational architectures, and complexity science to name a few.

Dr Jonathan Terry (IMNS)

JT1 Process and Device Simulation as an aid to Microelectronics Teaching (3 Projects)

Keywords: microelectronic devices, simulation, higher education teaching

The project student(s) will develop a laboratory to assist in the teaching of microelectronic device theory to undergraduate students. The Synopsys Sentaurus Workbench software platform will be used to develop simulation-based exercises to help UG students understand the theory and concepts of microelectronic devices and the effect that their fabrication can have on their performance. The following are suggested topic, however the project student can select their own area if is a suitable fit with the UG curriculum:

•          Short channel effects in MOSFET devices

•          Bipolar junction transistors

•          Microelectronic device building blocks

 The key objective will be to develop a complete learning experience to complement the lectures and other materials delivered to the UG cohort.

Dr Istvan Gyongy (IMNS)

IG1 Optical blood pressure monitoring using a single-photon counting sensor

Keywords: Signal processing, biomedical sensing, wearable technology

Recent years have seen a proliferation of wearable devices for measuring physiological parameters such as heart rate or blood oxygenation. However, continuous, cuff-less monitoring of blood pressure is still a challenge, especially in the form of a compact, portable instrument. One of the most promising techniques is by measuring the pulse transmit time (PTT), in other words the time for arterial pressure waves to reach a certain part of the body. PTT is found to be strongly correlated to blood pressure, although calibration is needed on individual subjects.

This project seeks to apply a single-photon sensitive (SPAD) image sensor, developed in the School of Engineering, to measure PTT by optical means. Optical fibres will be used to couple the image sensor, together with LED illumination sources, to two different locations on the body. Appropriate signal processing schemes will then be devised to analyse the reflected signals, as captured by the image sensor, and extract PTT.

IG2 MEMS scanning for single-photon LIDAR

Keywords: 3D imaging, MEMS, FPGA

3D time-of-flight/LIDAR devices are becoming a key component in many robotic and autonomous systems, such as self-driving cars. However, achieving a large field of view (FOV) in combination with a high frame rate can be difficult, impacting the perception of such systems in highly dynamic environments.

In view of this problem, the project will explore the use of advanced MEMS mirrors in scanned 3D-time of flight/LIDAR imaging using single-photon sensitive (SPAD) image sensors. An existing SPAD sensor, developed in the School of Engineering, will be used, in combination with an off-the-shelf MEMS mirror enabling point-to-point scanning. An FPGA will be programmed to control both the SPAD and the MEMS device. Both monostatic (the SPAD is coaxial with the LASER illumination) and bistatic (SPAD has wide angle optics and only the LASER is scanned) configurations will be considered, together with smart (scene-dependent) scanning schemes.

IG3 3D imaging using a flash, direct time-of-flight sensor

Keywords: 3D imaging, signal processing, neural networks

Smart phones and tablets with 3D imaging functionality typically use indirect time-of-flight (iToF) sensors based on continuous wave illumination, and a receiver with photo-demodulator pixels. However, recent devices from Apple and Samsung have adopted direct time-of-flight (dToF) arrays, where each pixel has the capability to detect and time individual photons, in combination with a pulsed light source. These arrays generate high accurate depth maps, which are unaffected by multi-path reflections, a common problem in iToF systems leading to significant inaccuracies in some scenes. The main downside of current dToF imagers is a limited array size, and thus lateral (x-y) resolution. This project will therefore consider the upscaling of dToF depth data based on high resolution intensity data. It will use a prototype dToF camera developed in the School of Engineering, together with a conventional RGB image sensor (as commonly found on mobile devices), with the aim being the generation of high resolution, colour 3D images.

Prof. Ian Underwood (IMNS) https://www.eng.ed.ac.uk/about/people/prof-ian-underwood

IU1 Projects related to 3D imaging (3 Projects)

Topics will be related to:

•          Optical sensing in 3D using

o   STMicro Flightsense sensors.

o   Intel Realsense Camera

o   Sensor integrated in smart phone

•          Visualising 3D data using

o   Head Mounted Displays

o   Light Field Displays

Dr. Filippo Menolascina (IBioE)

FM1 Deep-learning for segmentation of yeast cells

In this project you will use U-Net, a deep artificial neural network to automatically identify yeast cells in microscopy images of microfluidic devices. The aim of this project is to carry out a search in the space of hyper parameters of U-Net to minimise validation error. This project requires a basic understanding of matrix algebra and Python programming.

FM2 Deep-learning for segmentation of E. coli cells

In this project you will use U-Net, a deep artificial neural network to automatically identify bacterial cells in microscopy images of microfluidic devices. The aim of this project is to carry out a search in the space of hyper parameters of U-Net to minimise validation error. This project requires a basic understanding of matrix algebra and Python programming.  

Dr Jamie Marland (IMNS)

JM1 Optical tissue oxygenation monitoring

Keywords: Biomedical, optical sensing, microcontrollers, translational research

Oxygenation of tissues in the body is essential, but can be compromised following surgery in hospital due to localised failure of the blood supply. This can lead to dangerous complications for the patient. The aim of this project is to explore the use of miniature optical oxygen sensors (similar to those found in wearable fitness trackers) for implantation and post-surgical monitoring of tissue oxygenation. It will involve writing firmware for communication between the sensor and a microcontroller, and development of a visual interface for display of the data to clinicians. There will be opportunities for co-design of the display system with practicing NHS surgeons. This work will lay the foundations for a future trial of the technology in preclinical in vivo models.


on the way to be a Electrical Engineer & Designer......