MonBaby at BabyShow NYC 2018
MonBaby presented a new camera prototype and demo’ed MonBaby products.
Read More
A prototype for the complete baby monitoring solution – MonBaby3.0
We are excited to try a fresh from the mill prototype for the complete baby monitoring solution including audio, video and movement monitors combined into one stream of data. Please check out the video below:
Here is the graphical design for the MonBaby 3.0 enclosure:
Keep your eye peeled for more updates!
Read MoreUsing autocorrelation model for movement detection
We have noticed markable improvements in move detection in accelerometer signal when we tested autocorrelation model. It’s similar to autoregressive (AR) model, but in this article we call this model autocorrelation.
The autocorrelation is defined as:
Where N is the total window size, N win is the AR window size, i is the AR shift, and Ncor is the number of steps.
In tests we have noticed that autocorrelation works a lot better than other methods, including FFT., though it requires significant calibration and tuning. See the graphs with different move states below.
The final plot depicted below is the result of running movement detection using autocorrelation on live production data collected from one of our customers baby,
The three colored axis show row accelerometer signal and the dotted line shows detected movement in moves per minute. The bottom graph shows state enumeration with different movement classification states detected. A thing to note that no-move detection is corresponds to flag=-2, which is an alarming condition to alert parents. In the sample below, the alarm have not been detected.
Read MoreThe article in Entrepreneur magazine about MonBaby
https://www.entrepreneur.com/
Read More
Improving classifier
Today I tried a simple improvement to classifier. I took a previous algorithm and enhanced the classifier categorization logic. Previously I’ve required that the model leading coefficient (c[0]) should be more than noise threshold for every data point. This was my “goodness of fit” criteria. If the coefficient was too small for any of the sample points, the classifier would produce “no movement signal” output. This have lead to a very unstable output oscillating between “signal” and “no signal”.
Instead I changed the criteria to require that within analyzed interval majority of points should have coefficient above noise threshold. Then the whole interval is marked as “movement signal” and the output will be an averaged values.
Here is the code in python for aggregating classifier output:
outSig = [freqNA] * (len(freqEstimate)+Nlen)
# do one more pass to determine if range has majority of records with the signal or majority without signal
for ii in range(len(freqEstimate)):
freqCnt, freqVal = 0, -1.0
for jj in range(Nlen):
# if frequency estimate is more than zero, that means classifier decided that the signal is good
if ii+jj > len(freqEstimate) – 1: break
#print(‘freqEstimate:{}’.format(freqEstimate[ii+jj]))
if freqEstimate[ii+jj] and freqEstimate[ii+jj] > 0:
freqCnt += 1
freqVal += freqEstimate[ii+jj]
if freqCnt > Nlen/2+1:
freqVal /= freqCnt
outSig[ii+Nlen] = freqVal
Here are the results of the new classifier that is based on a function of model coefficients f(c0,c1,c2).
Compare that to a previous classifier output:
As you can see the output signal (in black) is a lot more stable.
For comparison, I’m also showing the output of the modified classifier that is based on NMSE, rather than coefficients:
Here is the output from another bigger data sample with classifier based on coefficients:
and same sample and NMSE classifier:
One can conclude that this “smoothing” aggregation technique works a lot better.
Read MoreInc magazine featured MonBaby product
Inc magazine: How Baby Safety Technology is Outgrowing Its Infancy
https://www.inc.com/molly-reynolds/how-baby-safety-technology-is-outgrowing-its-infan.html
Read More
Few moments of fame on CBS
CBS had a segment about CE Week in NYC and also included our product, MonBaby.
See us at 1:04 mark
http://www.wdtv.com/content/news/Consumer-Electronics-Week-had-a-gadget-for-everyone–435262293.html
Read MoreLaunching the MonZzz – shockingly smart snoring air wearable
We are starting a new product campaign on Indiegogo for the MonZzz – shockingly smart snoring aid wearable. It saves your relationships from snoring. Above is a video link, check it out, it’s pretty funny!
We also start a thunderclap.it campaign. It’s a free way to support the start of the campaign. In two weeks, thunderclap will simultaneosly send facebook and twitter messages. But for this we need to collect 250 votes. Please go here https://www.thunderclap.it/projects/59280-say-bye-to-snoring and click on facebook and twitter support buttons.
Read MoreMondevices @ CE week NYC 2017
Mondevices has exhibited at CE week in New York city. It was even more successful than last year, with more exhibitors, more visitors and more interest. We also got filmed by CBS, see below.
Here are some highlights in pictures:
CBS is filming our booth
Read More
Fox and Friends segment about the MonBaby
http://video.foxnews.
NY Baby Show – review 1
Blog: http://www.usjapanfam.
Read More
Mondevices went to NY baby show
Mondevices team went to NYC baby show, the biggest baby show in US. We were also picked for Fox and Friends baby product showcase. We got other media coverage, links to the videos will follow shortly.
Read MoreMeasurements of BLE signal indoors
Measurements of BLE signal strength (RSSI) dependency on the range indoors
Introduction
There is not much results of Bluetooth Low Energy range measurements to be found on the web. The primary reason is that it’s hard to control for environment, wall and floor materials. That being said, the numerous use cases for Bluetooth LE, from now on BLE, exactly in that type of environment, indoors. We, at Mondevices, decided to fill that gap.
Description of Setup
We took various BLE devices that transmit packets about 10 times per second and varied antenna designs. Most of PCBs were based on TI chip CC254x family. We have also included a combination of CC2541 with a power amplifier CC2590 and also used Texas instrument RF Studio with a mounted antenna. Below is a detailed description of the devices used.
- CC2590 Wire = CC2541/CC2590 with 30.6mm wire antenna perpendicular to the PCB, TX power set to +10dBm see attachment CC2590_WIRE.jpg
- CC2590_MIFA = CC2541/CC2590 with MIFA wire antenna, TX power set to +10dBm, see attachment CC2590_MIFA
- TI_DEV = Texas Instrument evaluation module with CC2540, TX power set to +4dBm, see attachment TI_DEV.jpg
- DI_POLE2 = Your prototype di-pole antenna PCB, TX power set to +4dBm, see attachment DI_POLE2.jpg
- MollyCHIP1 = Standard monBaby PCB with standard chip antenna, TX power set to +4dBm, see attachment Standard_monBaby.jpg
- WIRE = Standard monBaby PCB with 30.6mm WIRE antenna perpendicular to PCB, TX power set to +4dBm, see attachment monBaby_WIRE.jpg
Name | UUID | Description |
CC2590 MIFA | 68:9e:19:18:7a:a0 | CC2541+CC2590 amplifier using MIFA antenna |
WIRE | 20:91:48:4c:9b:c5 | CC2540 with vertical 1/4 wave-length antenna |
NoBattery | b4:99:4c:74:d6:99 | CC2540 with chip antenna and battery removed |
CC2590 Wire | 68:9e:19:18:83:f7 | CC2541+CC2590 amplifier using wire antenna |
TiDev | b4:99:4c:64:69:26 | TI Smart RF development kit antenna |
Roach1 | b4:99:4c:74:e8:c5 | CC2540 with wire antenna bent like roach antenna |
Wire3Coil | b4:99:4c:74:c5:92 | CC2540 with wire coiled around circular enclosure |
NoBattery | b4:99:4c:74:d6:99 | CC2540 with battery far away from PCB |
Molex Antenna | b4:99:4c:74:e4:8d | CC2540 with Molex spiral antenna |
Roach2 | b4:99:4c:74:b7:d5 | Roach 2 has antenna bent close to PCB |
MIFA | b4:99:4c:74:ef:0a | CC2540 (inverse F) MIFA antenna in the enclosure |
CHIP2Louise | b4:99:4c:74:b2:89 | CC2540 chip antenna called Louise in the enclosure |
CHIP1Molly | 20:91:48:4c:a3:8a | CC2540 chip antenna called Molly in the enclosure |
We have tried to find the best performing devices that could operate at furthers distances from the office and keep steady connection. We have found that TI RF Dev Kit and CC2590 booster are consistently perform better than other devices. TI Dev Kit has a large wave-length antenna and CC2590 boosts power transmission to 10 dbm, which is 4 times more than current chip antenna at 4 dbm setting or 8-9 times better than CC2541 0 dbm setting.
The photo of BLE range test setup describes positioning of all the test devices.
We tried all different combinations of antenna while trying to stay as close to same technology stack of TI CC254x family.
The close up of PCB with 1/4 wire antenna, CC2590 combo with MIFA and wire antenna.
Measurements
We have conducted measurement of RSSI strength of advertising packets of all 13 devices using TI Smart RF packet sniffer. We stored the results and parsed them with our custom parser script written in python. The results are attached as plots and tables.
Results
We did several measurements starting within in the same room in the office confines, then outside in the most RF challenging area of the kitchen and outside in the long corridor. The results of RSSI strength measurements of advertising packets are shown below. To avoid clatter on graphs we only display top three and top bottom plots. The tables shows stats for all
the devices. The kitchen area is very typical to every office and home, with metal surfaces of refrigerator, fax and printer machines and surrounding offices with BLE speakers, wearables and printers. Metal surfaces create destructive interference, where BLE signal interfere with itself. Other electronic devices create electromagnetic RF interferences. We counted around 200 BLE devices sending advertisement packets.
In the same room
In the office, same room where we placed all the devices, we can detect all the signals. This is an expected and obvious result. What is interesting that the biggest RSSI values we get from TI Dev Kit, CC2590 booster combo and, most surprising, just a simple vertical 1/4 wave length antenna positioned perpendicular to the PCB plane
The table below shows the stats for all the devices, not just top and bottom three. As one can see devices with power booster, CC2590, shows consistently on top.
UUID | Alias | MeanRSSI | StdDevRSSI | VarianceRSSI | MinRSSI | MaxRSSI | Count | FCSerrors | FCSerror% |
68:9e:19:18:83:f7 | CC2590 Wire |
-37.58 |
2.86 |
8.16 |
-50 |
-33 |
187 |
0 |
0 |
68:9e:19:18:7a:a0 | CC2590 MIFA |
-34.21 |
3.45 |
11.9 |
-47 |
-30 |
187 |
0 |
0 |
b4:99:4c:74:e8:c5 | Roach1 |
-44.31 |
2.95 |
8.72 |
-55 |
-40 |
186 |
0 |
0 |
20:91:48:4c:a3:8a | CHIP1Molly |
-55.22 |
2.98 |
8.87 |
-66 |
-51 |
186 |
2 |
1.08 |
b4:99:4c:64:69:26 | TiDev |
-42.95 |
3.27 |
10.71 |
-54 |
-38 |
185 |
0 |
0 |
20:91:48:4c:9b:c5 | WIRE |
-42.48 |
3.34 |
11.17 |
-54 |
-36 |
185 |
0 |
0 |
b4:99:4c:74:b7:d5 | Roach2 |
-57.35 |
3.13 |
9.78 |
-68 |
-50 |
176 |
6 |
3.41 |
b4:99:4c:74:e4:8d | Molex Ant |
-58.27 |
3.07 |
9.45 |
-68 |
-51 |
172 |
14 |
8.14 |
b4:99:4c:74:c5:92 | Wire3Coil |
-59.29 |
2.82 |
7.97 |
-67 |
-53 |
156 |
10 |
6.41 |
b4:99:4c:74:b2:89 | CHIP2Louise |
-61.53 |
2.4 |
5.78 |
-67 |
-58 |
97 |
28 |
28.87 |
b4:99:4c:74:d6:99 | NoBattery |
-60.29 |
2.82 |
7.95 |
-66 |
-54 |
78 |
18 |
23.08 |
b4:99:4c:74:ef:0a | MIFA |
-64.29 |
1.96 |
3.82 |
-68 |
-60 |
21 |
17 |
80.95 |
As one can see the strongest signals are from CC2590 combo with and without WIRE antenna, and from CC2540 with WIRE antenna. The percentage of FCS errors that demonstrate the data loss is also the least in CC2590+CC2541 combo and in CC2540 with the WIRE antenna.
In the RF dead zone – the kitchen
The kitchen is one of the most challenging from RF interference perspective environments. Here only a few devices survived.
The table below shows the stats for all the devices in the kitchen.
UUID | Alias | MeanRSSI | StdDevRSSI | VarianceRSSI | MinRSSI | MaxRSSI | Count | FCSerrors | FCSerror% |
68:9e:19:18:83:f7 | CC2590 Wire |
-60.59 |
1.87 |
3.5 |
-66 |
-57 |
133 |
0 |
0 |
20:91:48:4c:9b:c5 | WIRE |
-64.62 |
1.7 |
2.87 |
-70 |
-58 |
125 |
6 |
4.8 |
b4:99:4c:74:e8:c5 | Roach1 |
-68.07 |
1.45 |
2.1 |
-73 |
-65 |
114 |
10 |
8.77 |
68:9e:19:18:7a:a0 | CC2590 MIFA |
-68.44 |
2.94 |
8.63 |
-76 |
-62 |
114 |
14 |
12.28 |
b4:99:4c:74:b7:d5 | Roach2 |
-74.67 |
1.28 |
1.65 |
-77 |
-72 |
63 |
33 |
52.38 |
b4:99:4c:64:69:26 | TiDev |
-75 |
1.68 |
2.83 |
-79 |
-71 |
48 |
21 |
43.75 |
b4:99:4c:74:e4:8d | Molex Ant |
-75.98 |
1.01 |
1.02 |
-79 |
-73 |
46 |
34 |
73.91 |
b4:99:4c:74:c5:92 | Wire3Coil |
-76.94 |
1.48 |
2.18 |
-80 |
-75 |
32 |
25 |
78.13 |
b4:99:4c:74:d6:99 | NoBattery |
-77 |
2.6 |
6.77 |
-81 |
-71 |
13 |
7 |
53.85 |
b4:99:4c:74:b2:89 | CHIP2Louise |
-80 |
0 |
0 |
-80 |
-80 |
4 |
4 |
100 |
b4:99:4c:74:ef:0a | MIFA |
-79.67 |
0.47 |
0.22 |
-80 |
-79 |
3 |
3 |
100 |
20:91:48:4c:a3:8a | CHIP1Molly |
-80 |
1 |
1 |
-81 |
-79 |
2 |
1 |
50 |
Again in the kitchen are the strongest signal and lowest errors are in CC2590 combos and in WIRE antenna.
When we tried to walk beyond the kitchen table we only observed CC2590 combo surviving the distance and kitchen interference with high rate of data errors.
In the long corridor
We also did measurements in the corridor far away from the kitchen we got slightly better results.
UUID | Alias | MeanRSSI | StdDevRSSI | VarianceRSSI | MinRSSI | MaxRSSI | Count | FCSerrors | FCSerror% |
68:9e:19:18:7a:a0 | CC2590 MIFA |
-66.67 |
2.92 |
8.52 |
-78 |
-63 |
87 |
1 |
1.15 |
20:91:48:4c:9b:c5 | WIRE |
-71.19 |
2.89 |
8.33 |
-81 |
-68 |
68 |
7 |
10.29 |
b4:99:4c:74:d6:99 | NoBattery |
-74.62 |
1.74 |
3.04 |
-83 |
-72 |
60 |
13 |
21.67 |
68:9e:19:18:83:f7 | CC2590 Wire |
-73.67 |
3.28 |
10.76 |
-80 |
-67 |
52 |
7 |
13.46 |
b4:99:4c:64:69:26 | TiDev |
-72.45 |
3.06 |
9.34 |
-79 |
-67 |
42 |
4 |
9.52 |
b4:99:4c:74:e8:c5 | Roach1 |
-75.35 |
1.85 |
3.4 |
-78 |
-71 |
34 |
9 |
26.47 |
b4:99:4c:74:c5:92 | Wire3Coil |
-77.6 |
0.8 |
0.64 |
-79 |
-76 |
15 |
4 |
26.67 |
Not surprisingly the strongest signals and lowest FCS errors are from CC2590 combo with and without WIRE antenna, and from CC2540 with WIRE antenna.
Walking towards the advertising devices
The last test in this document we have walked from the area past the kitchen towards the advertising devices. Below is the graph that shows graphs and tables with RSSI values.
UUID | Alias | MeanRSSI | StdDevRSSI | VarianceRSSI | MinRSSI | MaxRSSI | Count | FCSerrors | FCSerror% |
68:9e:19:18:7a:a0 | CC2590 MIFA |
-58.8 |
10.72 |
115.02 |
-74 |
-34 |
51 |
2 |
3.92 |
68:9e:19:18:83:f7 | CC2590 Wire |
-58.66 |
11.73 |
137.62 |
-77 |
-34 |
50 |
5 |
10 |
20:91:48:4c:9b:c5 | WIRE |
-61.04 |
9.8 |
95.95 |
-76 |
-38 |
45 |
3 |
6.67 |
b4:99:4c:64:69:26 | TiDev |
-58.64 |
12 |
144.09 |
-77 |
-30 |
42 |
3 |
7.14 |
b4:99:4c:74:e8:c5 | Roach1 |
-59.68 |
9.36 |
87.57 |
-78 |
-43 |
31 |
2 |
6.45 |
b4:99:4c:74:d6:99 | NoBattery |
-61.81 |
12.61 |
159.04 |
-80 |
-42 |
27 |
3 |
11.11 |
b4:99:4c:74:e4:8d | Molex Ant |
-61.61 |
10.48 |
109.8 |
-82 |
-47 |
23 |
4 |
17.39 |
b4:99:4c:74:b7:d5 | Roach2 |
-62.95 |
8.07 |
65.09 |
-76 |
-51 |
21 |
1 |
4.76 |
b4:99:4c:74:c5:92 | Wire3Coil |
-64.68 |
7.5 |
56.22 |
-77 |
-50 |
19 |
4 |
21.05 |
b4:99:4c:74:ef:0a | MIFA |
-59.94 |
8.04 |
64.68 |
-74 |
-49 |
16 |
1 |
6.25 |
b4:99:4c:74:b2:89 | CHIP2Louise |
-64 |
4.55 |
20.67 |
-72 |
-58 |
12 |
2 |
16.67 |
20:91:48:4c:a3:8a | CHIP1Molly |
-63.5 |
6.95 |
48.25 |
-75 |
-52 |
12 |
5 |
41.67 |
Just like in previous experiments, we can see that the strongest signals and lowest data errors levels are in CC2590 combos.
Conclusions
We have conducted experiments indoors using different antenna configurations and designs on CC254x IoT SoC technology.
We have consistently observed that the strongest signals that survived both RF interference and distance are
- CC2590+CC2541 combo
- WIRE antenna
Read More
A blog about the MonBaby
Karen Lightman, Executive Director of MIG wrote a blog about the MonBaby technology-showcase at MIG Congress.
Here is a picture of Eugene, our COO with Karen
Read More
MonBaby is a finalist at MEMS Technology Showcase – MIG 2015 Congress
Mondevices went to MEMS Industry Group as a Technology Showcase finalist! The results will be announced in a few days, we are holding our fingers crossed.
We got a bunch of references in press for that:
Here is a snippet:
“MEMS Industry Group (MIG)’s annual MEMS & Sensors Technology Showcase atMEMS Executive Congress® US 2015 (November 4-6, 2015 in Napa, CA) highlights some of the newest and most compelling MEMS/sensors-based applications and technologies on the market. MIG today announced the shortlist of finalists who will compete for the title of winner at this year’s event.
MonBaby Movement and Rollover Monitor
Anxious parents can sleep more soundly now that MonBaby® is on the scene. An award-winning baby monitor that snaps like a button onto any article of a child’s clothing, MonBaby gives new parents peace of mind and helps them to sleep better knowing that they will receive an audible alarm on their smartphone if the baby rolls onto his or her stomach during sleep or stops movements.”
Read MoreMondevices at ABC Kids Expo – 2015
Mondevices demoed the MonBaby Smart Button monitor at the 2015 ABC Kids Expo in Las Vegas! Our team was on the ground at booth #2236 now through October 21.
ABC Kids is one of the nation’s largest expos for baby and children’s products. We’re excited to connect with journalists and retailers, and also to see some of the other wonderful new additions to the children’s products industry.
Check out the expo twitter updates: https://twitter.com/ABCKidsEXPO
Read More
What is the best technology solution for mobile Bluetooth(Smart)-WiFi bridge
Here is a comprehensive list of hardware available solutions:
BLE PAN
BeagleBone + WiFi-BLE: Page on boardzoo.com $49.99, requires a plastic enclosure, for comparison: BeagleBoard $55 Texas Instruments BeagleBone with SensorTag
Bluetooth mesh network
Internet of Things $49
Zuli Smartplug $55
BLE-WiFi bridge
What’s BluFi? The Bluetooth-to-WiFi Sensor – BluVision $89
WiFi BLE SoC
WL1271-TIWI-BLE | Wi-Fi (IEEE 802.11 (no longer supported by TI, still supported by LSR)
802.11 b/g/n WLAN Module $22.50
TI WiFi BLE combos:
WiLink Combo Solutions – Overview – TI.com
WiLink $39.72
CC3000($34.95) + CC2650
mBeacon: there is no price or any indication that actually an existing product, looks like 3D rendering
Beacon Hub – Bluetooth Low Energy Beacon to Wi-Fi Gateway Solution
TI 3200 launchpad: SimpleLink Wi-Fi CC3200 LaunchPad $30 (SimpleLink Wi-Fi Starter software)
McGateway
McThings: Tiny Wireless Bluetooth Sensors and Controlers $99
mcGateway – (Pre-order Only) – mcThings $80
Waspmote:
Bluetooth Low Energy to Connect Sensors with Smartphones and Tablets
Waspmote Starter Kit | M2M Connectivity $330 AUD (=$260)
CB-OLS BLE WLAN radio
cB OWL355 Multi-radio module
PRESS RELEASE: connectBlue® Announces Multiradio Module for Internet of Things
Recent Comments