公开(公告)号：US20160325505A1

公开(公告)日：2016-11-10

申请号：US15147629

申请日：2016-05-05

Applicant: Massachusetts Institute of Technology

Inventor： Jifei Ou ,  Gershon Dublon ,  Chin-Yi Cheng ,  Liang Zhou ,  Hiroshi Ishii

IPC: B29C67/00 ,  B65G27/16 ,  G06F3/043

Abstract: A computer produces a digital model that efficiently describes a dense array of many micro-pillars. The digital model achieves this efficiency by describing an entire micro-pillar with only a few parameters. For example, an entire micro-pillar may be described by two of more of the following parameters: height, base thickness, profile and tilt. The computer outputs instructions to fabricate the micro-pillar array, in accordance with the digital model. A 3D printer fabricates the micro-pillar array, based on the instructions. Applying vibration to a directional array of micro-pillars may cause the array of micro-pillars to actuate motion of a passive object that is touching the array. Also, a sensor may measure sounds caused by swipes against a micro-pillar array, and output signals indicative of the measurements. A computer performs a machine learning algorithm that takes the measurements as an input, and classifies the swipes.

Abstract translation: 计算机生成一个数字模型，可以有效地描述许多微支柱的密集阵列。 数字模型通过仅描述几个参数来描述整个微柱来实现这一效率。 例如，整个微柱可以通过以下两个参数来描述：高度，基底厚度，轮廓和倾斜度。 计算机根据数字模型输出制造微柱阵列的指令。 3D打印机根据说明制造微柱阵列。 对微柱的方向阵列施加振动可能导致微柱阵列致动与阵列接触的被动对象的运动。 此外，传感器可以测量由针对微柱阵列的滑动引起的声音，并输出指示测量的信号。 计算机执行以测量为输入的机器学习算法，并对滑动进行分类。

公开(公告)号：US10509559B2

公开(公告)日：2019-12-17

申请号：US15147629

申请日：2016-05-05

Applicant: Massachusetts Institute of Technology

Inventor： Jifei Ou ,  Gershon Dublon ,  Chin-Yi Cheng ,  Liang Zhou ,  Hiroshi Ishii

IPC: B29C64/386 ,  B33Y10/00 ,  B33Y50/02 ,  G06F3/0488 ,  B29C64/129 ,  B33Y30/00 ,  B33Y80/00

Abstract: A computer produces a digital model that efficiently describes a dense array of many micro-pillars. The digital model achieves this efficiency by describing an entire micro-pillar with only a few parameters. For example, an entire micro-pillar may be described by two of more of the following parameters: height, base thickness, profile and tilt. The computer outputs instructions to fabricate the micro-pillar array, in accordance with the digital model. A 3D printer fabricates the micro-pillar array, based on the instructions. Applying vibration to a directional array of micro-pillars may cause the array of micro-pillars to actuate motion of a passive object that is touching the array. Also, a sensor may measure sounds caused by swipes against a micro-pillar array, and output signals indicative of the measurements. A computer performs a machine learning algorithm that takes the measurements as an input, and classifies the swipes.