This project's funding goal was not reached on September 8, 2014.
About this project
This short video of the Hema-Imager 64x62 with Bluetooth shows the three primary image modes: thermal overlay on color camera, thermal overlay on wireframe camera, and raw thermal. Transparency, colormap selection and range are manually configurable.
How does the 64x62 Hema-Imager stack up against the competition? Check out this comparison table based on the specs:
Check out recent updates with images and video about Image interpolation (#5), first images with WiFi (#9), improvements in image quality (#16), and Erik's IAmA on Reddit, where he answered questions about thermal imaging. Please also check out our friends behind the IR-Blue: RH Workshop and MarDaSo. And finally our supplier Boston Electronics and the array manufacturer Heimann Sensor GMBH.
Lets talk about how thermal imaging can help us
Did you know that common building issues can be easily detected with thermal imaging before they become costly problems?
The joys of home ownership... Unexpected home repairs, costly energy bills and weather damage. Unless you are the first owner of a brand new home, most people have experienced a handful of issues year after year. In fact, on average, we spend $3,300 per year on home repairs and lose up to 30% of our heating or cooling on fixable problems! If you're like me, you have wondered if there is an affordable way to predict and isolate these issues before they get to be costly or damaging. Through research, I found out that thermal imaging is a great solution.
Thermal imaging is used by professional tradesman to detect issues that, if not fixed, could turn into very expensive problems. The trouble is that these devices are too costly for homeowners to justify the purchase (entry-level devices are typically between $1,000 and $5,000). Armed with this knowledge, I set off on a quest to create an affordable thermal imaging device that homeowners can use to find these problem areas in a home that are undetectible to the naked eye.
Our solution is the Hema-Imager, a heat mapping device that pairs with the camera from a smartphone or tablet and generates thermal images in real-time. The Hema-Imager (short for Heat Mapping Imager) allows homeowners, tradesmen, home inspectors, electrical service engineers and facilities managers to predict problem areas in equipment, homes, offices and other buildings by exposing unexpected temperatures.
Imagine that you have a 30 year-old light fixture and you are uncertain whether or not it was installed correctly. Wouldn't you want to know if heat was building up in the electrical fixture so that you could prevent a potential house fire? Thermal imaging shows you if something is hotter (or colder) than ambient temperatures, which is useful information when determining that something might be wrong. Thermal imaging can reveal dangerous load imbalances, bad electrical connections and other service equipment issues.
Thermal imaging is also great at detecting elevated levels of moisture in walls, floors and roofs. Wherever something is damper than the surrounding air, it will be cooled down several degrees by evaporation. Water damage in walls, very small leaks, even subtle moistness can be detected
Thermal imaging is particularly good at detecting residual water. In addition to homeowners, property inspectors and insurance adjustors could find the Hema-Imager invaluable for fast, affordable detection of water damage. Check out the video below. It show thermal can detect only a teaspoon of water squirted onto 1/2" drywall and allowed to drip off. Cooling was detectable on the other side of the panel after an hour.
A large portion of our energy goes to heating and cooling our air and heating water. Energy efficiency depends on the proper functioning and integrity of these systems. Sealing leaks, inadequate insulation, radiant equipment failure and improper venting can be quickly identified with thermal imaging. The EPA estimates the typical house could save 20% on heating and cooling costs by sealing and insulating ductwork alone.
Thermal imaging can save you money by identifying costly airflow issues, such as where energy gets wasted on cooling or heating your windows due to drapes that hang over ductwork vents. Fixed- and dynamic-range imaging modes let you better see heat flowing in ductwork and furnaces. The video below shows how thermal makes it easy to identify efficiency problems you wouldn't ever notice otherwise.
The biggest industrial use of thermal imaging is predictive maintenance, where it is used to help identify and predict equipment failures in order to prevent costly downtime at manufacturing facilities. However, we have found that some emergency services professionals also find the Hema-Imager to be a practical and quick solution.
We discovered that volunteer firefighters sometimes have to size up a fire call before calling for more equipment. The benefit of owning the Hema-Imager is that its affordable and it does not need constant recalibration while using it. Other devices near this price range require constant interruption to recalibrate the image, and the Hema-Imager does not have this problem, making it a practical solution.
"Erik contacted me last summer to ask for advice on his wireless thermal imager project. Since then we've talked about my experience with the IR-Blue project and the functionality he was aiming to get in his project. This summer, Erik shipped me one of his prototype thermal imagers and a Samsung 8 tablet to test out. I tested the imager with the tablet and the Android app looked good and really just needs a little polish and refinement for the most part. I tested it with my iPhone 5s and iPad mini, and the iOS app is a good start. Erik has done a good job with this and I'm looking forward to seeing his project get funded".
About the Hema-Imager
The Hema-Imager is a wireless thermopile-based imaging device that is not only compatible with nearly any smart device but has better temperature accuracy (2 deg C), wider temperature range (-4F - 450F) and better sensor angular resolution (0.61 angular degrees) than the closest affordable competition. Our apps can be used to display and store thermal images only or to display and store combined thermal and optical camera images, with adjustable thresholds and colormaps.
A few key elements of the Hema-Imager require further development (listed below in Development Timeline). The hardware, including electronics, case and firmware, is developed, with a manufacturing plan in place. A beta version of the Android and alpha version of the iOS application software have been developed and are being tested, but require further development for consumers.
- 64x62 thermopile array with integrated optics.
- Best sensor resolution at this price at 0.61 degree angular resolution
- No non-uniformity correction needed with thermopile technology
- Frame rate up to the ITAR-TASS regulations limit of 9 frames per second to any fully Bluetooth- or WiFi-capable device.
- Low power consumption and 850 mAhr battery provides up to 8 hours of continuous use without charging, or over a month if just using for 10 minutes daily.
- Android application for smartphone or tablet
- iOS application for iPhone and iPad
- Python & OpenCV application for windows & linux desktop
- App or button-driven laser pointer and online temperature display aligned to center of field of view with 2 deg C accuracy
- Thermal measurement range: -4F to 450F
- Use with any smart device (see Attachment and Thermal-to-camera calibration)
- Stores securely in an ultralow power mode, wakes up with a single button press
- Quick-connect with your smartphone or computer within seconds
- Store thermal images with various colormaps or transparent overlay on camera images
- Store video stream of either thermal or transparent overlay on camera video
- Can be used at up to 20m (Bluetooth) or 100m (WiFi) distance from your smartphone or computer
- Standby alarm mode: set it up on a scene, capture, then set an alarm on a selected feature and walk away
How can this device work with any smart device?
Attachment: We are developing an adjustable attachment that can fit any smartphone between 80 and 150mm in length, with or without a case, covering the majority of smartphones in use today, regardless of where the camera is located. If you instead want to use the Hema-Imager with a larger phone or tablet, or you want to be able to more quickly attach it, we have a second slide-on attachment option (the idea for this option came from a volunteer firefighter friend of ours). The quick-change slide-on attachment uses a low profile patch adhered to the device or the device case and a slider clip on the Hema-Imager that allows you to rapidly attach it. We will provide several patches and removable backings to allow you to attach the Hema-Imager to multiple devices or tablet, or even just stick it on a wall or other fixed point. The Hema-Imager will ship with both attachment options.
Thermal-to-camera calibration: Our software allows you to use the Hema-Imager with any smart device. The Hema-Imager software pairs the thermal image to the camera image using a simple, quick calibration procedure you only have to perform once per device you want to use it with. The only requirement is that the device is attached to the same place every time. If the device position has to change, for example if you get a new smartphone case, it is easy to re-do this camera calibration process. Both of our attachment options allow you to quickly and easily secure the imager in the same place on your smart device every time.
How does thermal imaging work?
The materials all around us give off a small amount of long-wave infrared (LWIR) light. The amount of this LWIR light given off is proportional to the temperature at the surface. So, a person running a fever will give off more of this "light" than a healthy person. Just like a regular cell phone camera, we can make an image of this light using some optics and an array of sensors. LWIR light is blocked by many materials (we use a silicon lens which lets the light through), the amount of light given off is pretty weak, and the die holding the sensors also gives off light proportional to the die temperature. The calibration process is used to provide a die subtraction image for any temperature the thermopile array is operating at.
How does Hema-Imager compare to other thermal imaging technology?
Hema-Imager has three main benefits over the closest competitors: no re-calibration, smart device compatibility and sensor specs.
Our greatest benefit to customers is that Hema-Imager is the only thermal imager in this price range that does not need constant recalibration. The major thermal imaging manufacturers all use microbolometer arrays to convert thermal energy into images, but these are very sensitive to offsets introduced by continuous heating of the sensing array. To keep the images from drifting, these need a mechanical shutter to constantly recalibrate the images while in use (the non-uniformity correction, or NUC - pronounced "nuke"). A NUC places a uniform reference target in front of the imager and generally takes 2-3 seconds, during which time the imager can't see beyond the shutter target. Automatic shutters are common in expensive cameras but greatly increase cost, and affordable cameras using microbolometer arrays require a manual shutter. The major manufacturers have invested a lot of money into foundries for making microbolometers and firmly believe in it. They believe people won't mind having to manually correct their imager during use. Our technology uses thermopile arrays and does not have this sensitivity to the nonuniformity problem.
The ability to work with nearly any smart device means Hema-Imager will not be obsolete if your smartphone or tablet dies. Simply pair it with your new device, perform the camera-to-thermal spatial calibration once and you'll be back in business within only a few minutes. Additionally, Hema-Imager is not necessarily physically tied to your device, and can be used up to 100 meters away from your smart device.
Finally, the thermopile array we are using is at the top of its class. It has a much wider temperature range than most microbolometers (more than double the range) and good spatial resolution and temperature accuracy.
About Hema-Imager Development
Two and a half years ago, I set out to build an array of single pixel sensor chips recently released by Texas Instruments in order to build a simple thermal camera to monitor my first child's health shortly after she was born. The first 8x8 pixel prototype was ready in the Fall of 2012. While building it, I kept thinking and learning about other uses for thermal imaging. After seeing the IR Blue kickstarter, I looked into thermopile arrays and found a new, higher resolution array was just coming on the market . Once I had a basic prototype incorporating this new array the following summer, it was time to think seriously and I got in touch with Andy Rawson, developer of the IR Blue device, who advised me to get the product more fully developed and find local partners to help with development and manufacturing.
In the summer of 2013, I partnered with Mehrdad Ramazanali, an experienced electronics product developer. In 2013, Hema Imaging LLC was formed and over the next several months we developed the hardware, firmware and software, culminating in the Hema-Imager. We are ready to go to production with our hardware, but our software requires further development and we need to meet a minimum order quantity for the thermopiles. Each production run of the thermopiles is custom, so small runs are very expensive. In order to make these affordable, the thermopile array has to be purchased in large quantities. We need your help making this happen.
About the Technology
Heimann Sensor GmbH is a developer and manufacturer of thermopile arrays for thermal sensing and other infrared sensors, and leads the world in thermopile array technology. The physicist managing the HTPA array product for Heimann Sensor has been assisting us in development of our imager based on their array. Boston Electronics is the U.S. distributor of the thermopile arrays and has provided early access to the 64x62 and 32x31 HTPA arrays for the development of the Hema-Imager.
We first reached out to Heimann Sensor and Boston Electronics in November 2012, and obtained our first thermopile arrays for development in Spring 2013. At SPIE Defense Sensing and Security 2013, we met with the physicist and project manager of the HTPA thermopile array, and refined our first prototype design with his help. In February of this year, Erik traveled to San Francisco for SPIE Photonics West 2014 and, with help from Boston Electronics and space in their booth, got early feedback on our first 32x31 prototypes. We met again at SPIE DSS 2014 and showed our prototypes off in the Boston Electronics booth and received a lot of interest from industry professionals from companies such as FLIR, Testo AG and many other people who were curious about our tablet and smartphone thermal imagers.
Hema-Imager Development and Delivery Timeline
Upon funding, we will place our larger main order for thermopiles with Boston Electronics. The production calibration chamber design is almost ready for manufacture, based on the prototype calibration cell we’ve machined and are using now for the factory 5-point temperature thermal calibration process. Each step in the manufacturing process will be ready to go shortly after we are funded.
Further development is needed for several key items we will complete during and after our campaign:
- WiFi hardware version requires more testing (product refinement)
- Completion of non-adhesive attachment design (product refinement)
- Injection mold tooling for case and attachments (product refinement)
- Perform a test production run for board, assembly, calibration, and packaging
- FCC/CE testing of test production run product (product testing)
- Application user interface development
The circuit boards, circuit assembly, test & calibration, the injection molding and final product assembly will all be done in northeast Ohio. A local testing facility will perform the FCC testing at 6 weeks after funding, using our first injection-molded cases and our test production run circuit boards.
The Hema-Imager will be made with two different data transfer options: Bluetooth and Wifi. Our original development focused on Bluetooth and Android and allows over 8 hours of battery life. However, before Android v4.2, the Bluetooth stack is less robust and crashes at higher frame rates. We have also been having problems with Bluetooth at KitKat Android (4.4) and are looking into a solution. WiFi is more robust and supports nearly all devices. Battery life is lower with WiFi. With our prototype, we've obtained over 3 hours continuous streaming with a low-power WiFi module that draws under 120mA.
We now know that most people want WiFi because it will work with more devices, but we don't know how many people want the extra-long battery life with Bluetooth, so we are going to try to offer them both. The circuit modifications are kept to only the wireless modules, necessitating two solder paste setups and a single reel change for the pick and place machine, which will allow us to produce both at minimal additional cost. We're keeping them the same price to keep things simple. For a higher frame rate with most devices and for any Apple device, we recommend Hema-Imager with WiFi for the data transfer.If you are a backer, you will receive an email within a week of end of our campaign asking for your choice of WiFi or Bluetooth, and one week later you will receive an email confirming your chosen configuration and delivery date.
We have tested the Hema-Imager over Bluetooth with the Android Razr Maxx, Samsung Galaxy SIII and Samsung Tab 8. We have tested the Hema-Imager over WiFi with the iPhone 4, iPhone 5, iPad Mini and iPad 2 (and several Android devices including the units tested with Bluetooth - the WiFi version should work with any device with WiFi capability).
Risks and challenges
The Hema-Imager hardware design is completed and has been designed for manufacturability. The lead time on the thermopile arrays is 12 weeks, but we have managed to reduce this for the first 50 units with special help from Heimann Sensor and Boston Electronics. We will place our primary order as soon as we meet our minimum of 800 units, so the project manager can begin production planning. We can increase this order any time until several weeks after our campaign end date. These remaining units will arrive 12 weeks after order. We have tested our circuit board reflow to our satisfaction and will perform our first test run shortly after funding. If necessary, we can perform a second test run with sufficient time before we need to make our first limited production run.
The tooling and injection molding house we have partnered with has given us a lead time of 5 weeks from initial order to delivery of final plastic parts, but they will begin their preparations as soon as we attain funding. The local electronics contract manufacturers we are evaluating have given a lead time of 4-8 weeks from order to delivery for full assembly, test and calibration. At this date, we have Android, iOS and standalone Windows applications, and we expect our final production-quality applications to be completed well in advance of our anticipated ship date. Applications will continue to be maintained and further developed after ship date, and will be easily updateable.
Despite these preparations, it is possible that a component we have ordered will not arrive in time for us to meet our ship date. If you are backing our project, you will receive a weekly status report on each outstanding development component and how we are addressing any delays and unforeseen problems.
In any project like this, it is critical to maintain regular communication and site-visits to manufacturing partners leading up and during product assembly, calibration and shipping. This is one of the greatest benefits of using local suppliers, and we intend to fully use this to our advantage in turning out a quality product on time.
Shipping will be done to the address you specify in a backer survey we will send out at the end of the project. We'll confirm your address in an email, just like those "confirm" emails you get when you setup a credit card or sign up to a mailing list. It will have two links in it, one to click on if the address is correct and confirmed, and a second link to click on if there is a mistake. If there are any errors, we'll contact you, but this is how we'll confirm your address before we ship.Learn about accountability on Kickstarter
If you reach your stretch goal to make the 82x62 array, will super-early backers receive the new sensor?
Yes, they will! We will have our 50 64x62 arrays arriving within a few weeks, so those devices will get made in our test run regardless of whether we reach our strech goal. We want to reward our early backers, and we plan on doing a a device swap when the 82x62 devices are ready in early to mid-December.
The Hema-Imager has a wider temperature range (-4 to 450F versus 32 to 212F for FLIR ONE) and does not require you to slide a reference target over the imager every so often. Also, it is less expensive ($250 versus $350), works with all WiFi- or Bluetooth-capable devices, and finally has longer battery life (WiFi version expected 4 hours).
The Bluetooth module can theoretically handle over 1Mbps, but actual throughput depends most on the Bluetooth stack in the kernel on the host (the phone's operating system). Android's Bluetooth stack wasn't very good before 4.2, but it was replaced starting with 4.2 and can handle higher data transfers (see comments for more details). To get to 9fps, the Bluetooth module has to operate at 576kbps. The WiFi throughput is much higher, primarily because it was always expected to handle a lot of data.
Bluetooth on iOS for Apple products unfortunately has a very low data transfer rate and cannot be used for imaging. The transfer rate with Bluetooth classic on an Apple product is about as low as the rate with BLE. For this reason, WiFi is necessary with Apple products.
- (35 days)