• Stars
    star
    1,102
  • Rank 40,508 (Top 0.9 %)
  • Language
    Python
  • License
    Apache License 2.0
  • Created about 4 years ago
  • Updated about 1 year ago

Reviews

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

Repository Details

An open-source, motorized, and modular microscope built using LEGO bricks, Arduino, Raspberry Pi and 3D printing.

An open-source MICROSCOPE built using LEGO bricks, 3D-printing, Arduino and Raspberry Pi

Animation

Images

Stop-motion

Key Features

  • Fully motorized: Camera angle, sample position, magnification and focus can be adjusted precisely using six stepper motors.
  • Modular: Stages and modules can be arranged in any configuration depending on the application.
  • Versatile: Uniform illumination guarantees high quality imaging for a variety of samples with reflective or transparent surfaces.
  • Wide magnification range: Samples with features from several centimeters to several micrometers can be imaged without changing the objective lens.
  • Low-cost: The whole assembly costs from $200 to $400 depending on the features and the vendors of the electronic components.
  • The microscope uses a Raspberry Pi mini-computer with an 8MP camera to capture images and videos. Stepper motors and the illumination are controlled using a circuit board comprising an Arduino microcontroller, six stepper motor drivers and a high-power LED driver. All functions can be controlled from a keyboard connected to the Raspberry Pi or a separate custom-built Arduino joystick connected to the mainboard. LEGO bricks are used to construct the main body of the microscope to achieve a modular and easy-to-assemble design concept.

Diagram

Assembly instructions

Video (YouTube) Instructions (PDF)
YouTube
Circuit assembly (PDF) Raspberry Pi HQ camera test
Circuit HQ-camera

Operation principle

The microscope has a simple operation principle based on changing the magnification and the focus by adjusting the relative distances between a camera, a single objective lens and a sample. Briefly, two linear stages with stepper motors are used to adjust these distances for a continuous and wide magnification range. Four additional stepper motors tilt the camera module and change the X-Y position and rotation of the sample. A uniform light source illuminates the sample either from an angle (reflected light) or from the bottom of the sample (transmitted light). The system can also be used as a digital water contact angle goniometer by taking cross-section images of droplets.

Operation

Assembly steps

3D printing

I assembled the main body of the microscope using individually-purchased LEGO bricks. Instead of using motors and gears from LEGO Technic, I designed custom actuators using FreeCAD software and printed them using my personal 3D printer. This approach not only lowered the cost of the microscope but also gave me some flexibility in the design and implementation of precise linear and rotary actuators. In principle, the whole structure could be 3D-printed without using any LEGO parts but that would be less modular and more time consuming.

All STL files for 3D printing and the original FreeCAD files for editing are available in this folder. You may need to install FreeCAD Gear workbench to edit the gears.

⚠️ A good quality printing depends on many factors. I optimized the designs after several iterations of printing. If the parts do not match well, some minor modification in the original design file (e.g. enlarging the holes matching to LEGO studs) or polishing/drilling may be required. More information on the printer and slicer settings is given in the PDF document (page #72).

3Dprinting

Electronics

⚠️ This part requires some basic knowledge on electronic circuit design and Arduino.

All design and Gerber files of printed circuit boards (PCBs) are available in this folder. I designed the PCBs using DesignSpark PCB software. The list of all components can be found in the PDF document (page #71)

There is also a separate instruction manual giving more details about the assembly of the circuit boards.

The operation of the electronics is straight-forward. I used two Arduino microcontrollers, one for the mainboard (essential) and one for the joystick controller (optional), to control the stepper motors and the LED illumination of the microscope. In principle, everything could be controlled directly from the Raspberry Pi without using any Arduino microcontroller but I decided to leave testing this option for another time.

💡 The joystick controller is optional because the stepper motors and the LED can also be controlled from a keyboard connected to the Raspberry Pi. But having a joystick is fun!

Electronics

I preferred Adafruit ItsyBitsy 32u4 5V but any Arduino board with enough number of I/O pins should work. If you want to use the joystick controller, the Arduino boards should support serial (UART) communication because they cannot communicate with each other via their USB ports.

The Arduino on the controller circuit reads the status of three thumb joysticks (for six stepper motors) and a potentiometer (for the LED intensity) via its analog inputs. The data is sent to the Arduino on the mainboard via UART (RX, TX) serial communication. An OLED display on the controller can be used to display useful information, like the PWM (pulse-width modulation) intensity of the LED. The Arduino on the mainboard receives and processes the incoming data and sends stepper motor signals to the corresponding stepper motor via its respective motor driver. Six motor drivers share the same signal pins but only one driver is activated at a time using the enable (EN) or the sleep (SLP) pin of the driver. This implementation requires only 10 I/O pins (4 signal + 6 enable) instead of 24 (4 signal x 6 motors). It also prevents the heating of the motors when they are idle and limits the current consumption by allowing only one motor running at a time. Alternatively, a motor driver with a built-in indexer with STEP/DIRECTION control can be used to reduce the number of I/O pins but such drivers are more expensive. The data is sent only if there is a change in the joystick position is detected to avoid continuous communication.

The mainboard can be powered from an external 5V wall charger or directly from the USB port of the Raspberry Pi. The latter also allows USB communication between the mainboard and the Raspberry Pi to control the stepper motors and the LED intensity from a keyboard. The intensity of the LED is controlled by PWM using a dedicated pin on the Arduino. I recommend keeping the LED off while the Raspberry Pi is booting and then gradually increasing the intensity. The LED driver and the high-power LED used in this project required more than 6V for a rated operation. I initially used an external 12V power adapter for the LED but later generated 12V directly from the USB port of the Raspberry Pi using a step-up DC-DC converter. In this case, the 5V power adapter should be able to supply enough current for the Raspberry Pi, stepper motors and the LED. I tested the system using a single 5V / 3A supply (original Raspberry Pi 4 power supply or a wall charger). Different configurations may require different power ratings. For example, if the display is also powered from the same source, then a more powerful 5V source would be needed. Here is a good one.

⚠️ The high-power LED gets hot after some time, be cautious. I used a standard aluminum heat sink to help cooling a little bit.

Uploading the Arduino code

  • Download the latest version of the Arduino IDE
  • Install the libraries (Sketch → Include Library → Manage Libraries):
    • AccelStepper: used to control the stepper motors (mainboard).
    • (Optional) Adafruit NeoPixel: if you want to have a status LED or nice color effects in the mainboard.
    • (Optional) Adafruit_GFX and Adafruit_SSD1306: if you want to have an OLED display in the joystick to show the intensity of the LED illumination.
  • Add the board by following these instructions.
  • Upload the code(s).

Arduino

Final assembly

Follow the detailed instructions given in the PDF document. Briefly:

  • It is important to fix the 32x32 LEGO baseplate to a rigid table or support for a good mechanical stability. Rubber dampers or an air cushion can be used to minimize vibrations.
  • Unfortunately it is not possible to create a public shopping basket in the LEGO shop, all parts need to be added one by one. I listed the LEGO parts in this document but I recommend buying extra bricks and plates in case you need to change something.
  • Some LEGO parts are needed to be glued permanently for better stability. I provided recommendations in the document based on my experience. In general, it is a good idea to glue a smaller LEGO piece (e.g. 2x2) to a larger one (e.g. 2x4) to have a stronger interlocking while preserving the advantage of LEGO bricks for modularity.
  • After assembling the microscope and connecting all the cables, boot the Raspberry Pi.
  • ⚠️ Do not forget to enable the camera from the Raspberry Pi configuration (Preferences → Raspberry Pi configuration → Interfaces → Camera Enabled → Reboot) as explained here.

Assembly

Python-Raspberry Pi

I wrote a simple program in Python 3 to control the microscope, modify camera settings and take photos and videos from keyboard. The code allows changing almost all camera settings using keyboard shortcuts. The speed and the direction of the stepper motors and the LED intensity can also be controlled from the keyboard independently from the joystick controller.

The code requires following dependencies:

  • EasyGUI: to generate simple message boxes and select the folder where the images are saved
  • pySerial: to communicate with the Arduino via USB
  • pynput: to monitor keyboard inputs

There are two important parameters in the program to be changed according to your configuration:

HighResolution:

  • True for (3280x2464), tested with Raspberry Pi 4.
  • False for (1920x1080), tested with Raspberry Pi 3 and Zero.

KeyboardControl:

  • True if the Arduino mainboard is connected to the Raspberry Pi via USB.
  • False if the mainboard is not connected and powered from an external 5V supply

Finally, launch the program from the terminal and enjoy the microscope:

python3 /home/pi/MicroscoPy.py 

Program

Basic operation modes

Modes

Media coverage

IEEE Spectrum: "Build A Sophisticated Microscope Using Lego, 3D Printing, Arduinos, and a Raspberry Pi"

IBM - Medium article: "IBM open sources $300 fully-functional LEGO®microscope design"

Futurism: "This Engineer Published Scientific Papers Using a Lego Microscope"

License

This project uses the Apache License Version 2.0 software license.

Separate third-party libraries used in Arduino and Python codes are licensed by their respective providers pursuant to their own separate licenses.

About

The microscope is not currently available for purchase in kit or built form.

Copyright 2021 Yuksel Temiz

More Repositories

1

sarama

Sarama is a Go library for Apache Kafka.
Go
10,858
star
2

plex

The package of IBM’s typeface, IBM Plex.
CSS
9,297
star
3

css-gridish

Automatically build your grid design’s CSS Grid code, CSS Flexbox fallback code, Sketch artboards, and Chrome extension.
CSS
2,253
star
4

openapi-to-graphql

Translate APIs described by OpenAPI Specifications (OAS) into GraphQL
TypeScript
1,594
star
5

Project_CodeNet

This repository is to support contributions for tools for the Project CodeNet dataset hosted in DAX
Python
1,485
star
6

fp-go

functional programming library for golang
Go
1,480
star
7

pytorch-seq2seq

An open source framework for seq2seq models in PyTorch.
Python
1,431
star
8

fhe-toolkit-linux

IBM Fully Homomorphic Encryption Toolkit For Linux. This toolkit is a Linux based Docker container that demonstrates computing on encrypted data without decrypting it! The toolkit ships with two demos including a fully encrypted Machine Learning inference with a Neural Network and a Privacy-Preserving key-value search.
C++
1,427
star
9

ibm.github.io

IBM Open Source at GitHub
JavaScript
1,106
star
10

Dromedary

Dromedary: towards helpful, ethical and reliable LLMs.
Python
1,059
star
11

MAX-Image-Resolution-Enhancer

Upscale an image by a factor of 4, while generating photo-realistic details.
Python
863
star
12

elasticsearch-spark-recommender

Use Jupyter Notebooks to demonstrate how to build a Recommender with Apache Spark & Elasticsearch
Jupyter Notebook
806
star
13

differential-privacy-library

Diffprivlib: The IBM Differential Privacy Library
Python
774
star
14

build-blockchain-insurance-app

Sample insurance application using Hyperledger Fabric
JavaScript
719
star
15

FfDL

Fabric for Deep Learning (FfDL, pronounced fiddle) is a Deep Learning Platform offering TensorFlow, Caffe, PyTorch etc. as a Service on Kubernetes
Go
676
star
16

spring-boot-microservices-on-kubernetes

In this code we demonstrate how a simple Spring Boot application can be deployed on top of Kubernetes. This application, Office Space, mimicks the fictitious app idea from Michael Bolton in the movie "Office Space".
JavaScript
548
star
17

cloud-native-starter

Cloud Native Starter for Java/Jakarta EE based Microservices on Kubernetes and Istio
Shell
517
star
18

federated-learning-lib

A library for federated learning (a distributed machine learning process) in an enterprise environment.
Python
480
star
19

nicedoc.io

pretty README as service.
JavaScript
473
star
20

clai

Command Line Artificial Intelligence or CLAI is an open-sourced project from IBM Research aimed to bring the power of AI to the command line interface.
Python
466
star
21

import-tracker

Python utility for tracking third party dependencies within a library
Python
458
star
22

mac-ibm-enrollment-app

The Mac@IBM enrollment app makes setting up macOS with Jamf Pro more intuitive for users and easier for IT. The application offers IT admins the ability to gather additional information about their users during setup, allows users to customize their enrollment by selecting apps or bundles of apps to install during setup, and provides users with next steps when enrollment is complete.
Swift
454
star
23

mobx-react-router

Keep your MobX state in sync with react-router
JavaScript
437
star
24

openapi-validator

Configurable and extensible validator/linter for OpenAPI documents
JavaScript
429
star
25

EvolveGCN

Code for EvolveGCN: Evolving Graph Convolutional Networks for Dynamic Graphs
Python
384
star
26

fhe-toolkit-macos

IBM Homomorphic Encryption Toolkit For MacOS
C++
356
star
27

AutoMLPipeline.jl

A package that makes it trivial to create and evaluate machine learning pipeline architectures.
HTML
345
star
28

graphql-query-generator

Randomly generates GraphQL queries from a GraphQL schema
TypeScript
334
star
29

portieris

A Kubernetes Admission Controller for verifying image trust.
Go
329
star
30

BluePic

WARNING: This repository is no longer maintained ⚠️ This repository will not be updated. The repository will be kept available in read-only mode.
Swift
325
star
31

FedMA

Code for Federated Learning with Matched Averaging, ICLR 2020.
Python
320
star
32

lale

Library for Semi-Automated Data Science
Python
320
star
33

powerai-counting-cars

Run a Jupyter Notebook to detect, track, and count cars in a video using Maximo Visual Insights (formerly PowerAI Vision) and OpenCV
Jupyter Notebook
317
star
34

evote

A voting application that leverages Hyperledger Fabric and the IBM Blockchain Platform to record and tally ballots.
JavaScript
316
star
35

aihwkit

IBM Analog Hardware Acceleration Kit
Jupyter Notebook
314
star
36

zshot

Zero and Few shot named entity & relationships recognition
Python
308
star
37

blockchain-network-on-kubernetes

Demonstrates the steps involved in setting up your business network on Hyperledger Fabric using Kubernetes APIs on IBM Cloud Kubernetes Service.
Shell
305
star
38

IBM-Z-zOS

The helpful and handy location for finding and sharing z/OS files, which are not included in the product.
REXX
296
star
39

charts

The IBM/charts repository provides helm charts for IBM and Third Party middleware.
Smarty
295
star
40

TabFormer

Code & Data for "Tabular Transformers for Modeling Multivariate Time Series" (ICASSP, 2021)
Python
295
star
41

blockchain-application-using-fabric-java-sdk

Create and Deploy a Blockchain Network using Hyperledger Fabric SDK Java
Java
292
star
42

mac-ibm-notifications

macOS agent used to display custom notifications and alerts to the end user.
Swift
289
star
43

MAX-Object-Detector

Localize and identify multiple objects in a single image.
Python
286
star
44

design-kit

The IBM Design kit is a collection of tools aimed to help you design and prototype experiences faster, with confidence and thoughtfulness. This kit is based on the IBM Design System. Also, you may use this documentation to create add-on libraries to the IBM Design System or submit bugs to the current system.
272
star
45

AccDNN

A compiler from AI model to RTL (Verilog) accelerator in FPGA hardware with auto design space exploration.
Verilog
270
star
46

deploy-ibm-cloud-private

Instructions and Code required to install IBM Cloud Private
HCL
263
star
47

vue-a11y-calendar

Accessible, internationalized Vue calendar
JavaScript
253
star
48

audit-ci

Audit NPM, Yarn, and PNPM dependencies in continuous integration environments, preventing integration if vulnerabilities are found at or above a configurable threshold while ignoring allowlisted advisories
TypeScript
253
star
49

watson-banking-chatbot

A chatbot for banking that uses the Watson Assistant, Discovery, Natural Language Understanding and Tone Analyzer services.
JavaScript
250
star
50

UQ360

Uncertainty Quantification 360 (UQ360) is an extensible open-source toolkit that can help you estimate, communicate and use uncertainty in machine learning model predictions.
Python
249
star
51

Kubernetes-container-service-GitLab-sample

This code shows how a common multi-component GitLab can be deployed on Kubernetes cluster. Each component (NGINX, Ruby on Rails, Redis, PostgreSQL, and more) runs in a separate container or group of containers.
Shell
243
star
52

tensorflow-hangul-recognition

Handwritten Korean Character Recognition with TensorFlow and Android
Python
232
star
53

transition-amr-parser

SoTA Abstract Meaning Representation (AMR) parsing with word-node alignments in Pytorch. Includes checkpoints and other tools such as statistical significance Smatch.
Python
229
star
54

BlockchainNetwork-CompositeJourney

Part 1 in a series of patterns showing the building blocks of a Blockchain application
Shell
227
star
55

pytorchpipe

PyTorchPipe (PTP) is a component-oriented framework for rapid prototyping and training of computational pipelines combining vision and language
Python
223
star
56

Graph2Seq

Graph2Seq is a simple code for building a graph-encoder and sequence-decoder for NLP and other AI/ML/DL tasks.
Python
219
star
57

LNN

A `Neural = Symbolic` framework for sound and complete weighted real-value logic
Python
214
star
58

Scalable-WordPress-deployment-on-Kubernetes

This code showcases the full power of Kubernetes clusters and shows how can we deploy the world's most popular website framework on top of world's most popular container orchestration platform.
Shell
214
star
59

janusgraph-utils

Develop a graph database app using JanusGraph
Java
204
star
60

ModuleFormer

ModuleFormer is a MoE-based architecture that includes two different types of experts: stick-breaking attention heads and feedforward experts. We released a collection of ModuleFormer-based Language Models (MoLM) ranging in scale from 4 billion to 8 billion parameters.
Python
203
star
61

ibm-generative-ai

IBM-Generative-AI is a Python library built on IBM's large language model REST interface to seamlessly integrate and extend this service in Python programs.
Python
202
star
62

tensorflow-large-model-support

Large Model Support in Tensorflow
199
star
63

Scalable-Cassandra-deployment-on-Kubernetes

In this code we provide a full roadmap the deployment of a multi-node scalable Cassandra cluster on Kubernetes. Cassandra understands that it is running within a cluster manager, and uses this cluster management infrastructure to help implement the application. Kubernetes concepts like Replication Controller, StatefulSets etc. are leveraged to deploy either non-persistent or persistent Cassandra clusters on Kubernetes cluster.
Shell
195
star
64

adaptive-federated-learning

Code for paper "Adaptive Federated Learning in Resource Constrained Edge Computing Systems"
Python
193
star
65

action-recognition-pytorch

This is the pytorch implementation of some representative action recognition approaches including I3D, S3D, TSN and TAM.
Python
193
star
66

gantt-chart

IBM Gantt Chart Component, integrable in Vanilla, jQuery, or React Framework.
JavaScript
193
star
67

api-samples

Samples code that uses QRadar API's
Python
192
star
68

cdfsl-benchmark

(ECCV 2020) Cross-Domain Few-Shot Learning Benchmarking System
Python
190
star
69

kube101

Kubernetes 101 workshop (https://ibm.github.io/kube101/)
Shell
184
star
70

CrossViT

Official implementation of CrossViT. https://arxiv.org/abs/2103.14899
Python
180
star
71

browser-functions

A lightweight serverless platform that uses Web Browsers as execution engines
JavaScript
180
star
72

pwa-lit-template

A template for building Progressive Web Applications using Lit and Vaadin Router.
TypeScript
176
star
73

rl-testbed-for-energyplus

Reinforcement Learning Testbed for Power Consumption Optimization using EnergyPlus
Python
170
star
74

AMLSim

The AMLSim project is intended to provide a multi-agent based simulator that generates synthetic banking transaction data together with a set of known money laundering patterns - mainly for the purpose of testing machine learning models and graph algorithms. We welcome you to enhance this effort since the data set related to money laundering is critical to advance detection capabilities of money laundering activities.
Python
170
star
75

socket-io

A Socket.IO client for C#
C#
169
star
76

tfjs-web-app

A TensorFlow.js Progressive Web App for Offline Visual Recognition
JavaScript
164
star
77

molformer

Repository for MolFormer
Jupyter Notebook
163
star
78

spark-tpc-ds-performance-test

Use the TPC-DS benchmark to test Spark SQL performance
TSQL
160
star
79

watson-online-store

Learn how to use Watson Assistant and Watson Discovery. This application demonstrates a simple abstraction of a chatbot interacting with a Cloudant NoSQL database, using a Slack UI.
HTML
156
star
80

istio101

Istio 101 workshop (https://ibm.github.io/istio101/)
Shell
154
star
81

Medical-Blockchain

A healthcare data management platform built on blockchain that stores medical data off-chain
Vue
150
star
82

watson-assistant-slots-intro

A Chatbot for ordering a pizza that demonstrates how using the IBM Watson Assistant Slots feature, one can fill out an order, form, or profile.
JavaScript
143
star
83

tsfm

Foundation Models for Time Series
Jupyter Notebook
143
star
84

simulai

A toolkit with data-driven pipelines for physics-informed machine learning.
Python
142
star
85

etcd-java

Alternative etcd3 java client
Java
141
star
86

deploy-react-kubernetes

Built for developers who are interested in learning how to deploy a React application on Kubernetes, this pattern uses the React and Redux framework and calls the OMDb API to look up movie information based on user input. This pattern can be built and run on both Docker and Kubernetes.
JavaScript
139
star
87

innovate-digital-bank

This repository contains instructions to build a digital bank composed of a set of microservices that communicate with each other. Using Nodejs, Express, MongoDB and deployed to a Kubernetes cluster on IBM Cloud.
JavaScript
137
star
88

ipfs-social-proof

IPFS Social Proof: A decentralized identity and social proof system
JavaScript
135
star
89

KubeflowDojo

Repository to hold code, instructions, demos and pointers to presentation assets for Kubeflow Dojo
Jupyter Notebook
132
star
90

probabilistic-federated-neural-matching

Bayesian Nonparametric Federated Learning of Neural Networks
Python
132
star
91

fhe-toolkit-ios

IBM Fully Homomorphic Encryption Toolkit For iOS
C++
131
star
92

pytorch-large-model-support

Large Model Support in PyTorch
130
star
93

taxinomitis

Source code for Machine Learning for Kids site
JavaScript
127
star
94

Decentralized-Energy-Composer

WARNING: This repository is no longer maintained ⚠️ We are no longer showing the Hyperledger Composer Service.
TypeScript
127
star
95

quantum-careers

Learn about career opportunities with IBM Quantum.
126
star
96

cloud-pak

IBM Cloud Paks are enterprise-grade containerized software by combining container images with enterprise capabilities for deployment in production use cases with integrations for management and lifecycle operations. Features such as pre-configured deployments based on product expertise, rolling upgrades, and management of production workloads.
Shell
126
star
97

build-knowledge-base-with-domain-specific-documents

Create a knowledge base using domain specific documents and the mammoth python library
Jupyter Notebook
125
star
98

japan-technology

IBM Related Japanese technical documents - Code Patterns, Learning Path, Tutorials, etc.
Jupyter Notebook
125
star
99

DiffuseKronA

DiffuseKronA: A Parameter Efficient Fine-tuning Method for Personalized Diffusion Models
125
star
100

compliance-trestle

An opinionated tooling platform for managing compliance as code, using continuous integration and NIST's OSCAL standard.
Python
124
star