Differences

This shows you the differences between two versions of the page.

--- emrp:ws2025:agv [2026/02/28 22:14] – 23553_students.hsrw
+++ emrp:ws2025:agv [2026/02/28 22:52] (current) – [6. Discussion] 23553_students.hsrw
@@ Line 498: / Line 498: @@
     * keep heating when vent starts keep fans on
   * Humidity rise scenario
+<imgcaption image14|>
+{{ :emrp:ws2025:agvtestingsetup.jpg?&800 |Test setup 1. Source: Own image.}}
+</imgcaption>
 ==== 5.2 Data collection and logging ====
@@ Line 521: / Line 525: @@
 The average time across all runs was 4:07 min, with a slight increasing trend over the repetitions (later runs took longer).
-<imgcaption image14|>
+<imgcaption image15|>
 {{ :emrp:ws2025:agvtest1.jpg?&800 |Data of temperature test 1. Source: Own illustration.}}
 </imgcaption>
@@ Line 537: / Line 541: @@
 All three runs produced the same result, giving an average time of 3:50 min for this scenario.
-<imgcaption image15|>
+<imgcaption image16|>
 {{ :emrp:ws2025:agvtest2.jpg?&800 |Data of temperature test 2. Source: Own illustration.}}
 </imgcaption>
@@ Line 545: / Line 549: @@
 ==Humidity rise scenario:==
-In this scenario, humidity was increased using hot water/steam until the automation triggered ventilation at 70% RH. Indoor humidity peaked at 88% RH and was then reduced to 63% RH while the fans were running. The fans stopped automatically after the humidity dropped back below the threshold. After the fans stopped, indoor humidity began to rise again. When the fans were started manually, humidity continued to decrease until the inside and outside humidity levels were approximately equal. During the final part of the test, the outside humidity increased.
+In this scenario, humidity was increased using hot water/steam until the automation triggered ventilation at 70% RH. Indoor humidity peaked at 88% RH and was then reduced to 63% RH while the fans were running. The fans stopped automatically after the humidity dropped back below the threshold (Fig 17. - 1.). After the fans stopped, indoor humidity began to rise again (Fig 17. - 2.). When the fans were started manually, humidity continued to decrease until the inside and outside humidity levels were approximately equal. During the final part of the test, the outside humidity increased (Fig 17. - 3.).
-<imgcaption image16|>
+<imgcaption image17|>
 {{ :emrp:ws2025:agvtest3.jpg?&800 |Data of humidity test. Source: Own illustration.}}
 </imgcaption>
-===== 6. Results & Discussion =====
+===== 6. Discussion =====
+The system successfully measured indoor/outdoor temperature and humidity and controlled ventilation automatically via Home Assistant and ESPHome. In the controlled bathroom tests, the automation triggered reliably at the defined thresholds and ventilation reduced both temperature and humidity when outside conditions were more favorable than inside.
+One concern was the limited absolute accuracy of the DHT11 sensors, especially for relative humidity. However, the control logic primarily relies on relative comparisons between inside and outside rather than perfectly accurate absolute values. Using two identical sensor modules also helps because systematic offsets tend to cancel out when comparing trends and differences. As a result, the DHT11 accuracy is acceptable for this prototype, while higher accuracy sensors would improve confidence in the exact trigger thresholds.
+In the temperature baseline test (heater OFF, fans OFF), cooling back below the ventilation threshold took 4:07 min on average. With ventilation enabled after switching the heater off (heater OFF, fans ON), cooling took 3:50 min across all runs. With a 2000 W heater kept ON, ventilation could not restore temperature below the threshold, but it likely slowed the temperature increase. The baseline cooldown duration increased over repeated trials (from 3:50 to 4:20), which is plausibly explained by heat storage in the room’s surfaces (walls, tiles, furniture) and their gradual release after the heater was switched off. The ventilation-enabled cooling test was performed after the baseline series, meaning it started under less favorable conditions; despite this, the cooldown time with fans remained lower (3:50), supporting a real cooling effect from ventilation in this setup.
-==== 6.1 Results ====
+In the humidity test, ventilation started automatically at 70% RH. Indoor humidity peaked at 88% RH and was reduced to 63% RH while the fans were running. After the fans stopped, indoor humidity rose again even though no additional moisture was added, suggesting that without continued airflow some humid air pockets can remain and moisture on surfaces can re-equilibrate with the air. If humidity had climbed above 70% RH again, the automation would have restarted ventilation. For this test, the fans were started manually instead, and indoor humidity continued to decrease until it approached equilibrium with the outside sensor reading. Towards the end, the measured outside humidity also increased slightly, which is most likely an artifact of the indoor test setup with limited fresh-air exchange rather than a realistic outdoor effect.
+===== 7. Conclusion and outlook =====
-==== 6.2 Limitations ====
+This project demonstrates a working prototype of a Home Assistant–based greenhouse ventilation system using low-cost sensors and an ESP32 controller. The implementation covers the complete chain from sensing (inside/outside temperature and humidity) to automated actuation (relay-controlled intake/exhaust fans) and provides a practical foundation for further improvements. Initial indoor validation showed that the automation triggers reliably and that ventilation can measurably improve temperature and humidity conditions when outside air is more favorable than inside air.
-The current evaluation has several limitations:
+==== Outlook ====
-  * Testing was performed in a bathroom rather than a real greenhouse, so airflow patterns, heat capacity, and leakage behavior differ significantly.
-  * The controlled tests mainly covered cases where the “inside” environment was hotter and/or more humid than the “outside” reference. Conditions such as rain events, strong solar radiation, and rapid outside fluctuations were not fully represented.
-  * DHT11 sensors are low accuracy, especially for humidity (often ±5% RH or worse with a range of 20-80%).
-===== 7. Future Work Ideas =====
+The next step is the deployment and long-term validation in the real greenhouse once construction is completed. Beyond installation, several extensions could improve functionality, robustness, and autonomy:
-Here are a few ideas for future development, other than the actual deployment in the greenhouse.:
+  * Additional sensing
-  * Add additional sensors
     * Soil moisture sensing
     * Light and CO₂ sensing for improved climate control decisions
-  * Extend automation capabilities
+  * Extended automation features
     * Automatic watering based on soil moisture and schedules
-    * Artificial light
+    * Artificial lighting for plant support during low-light periods
-    * Absolute humidity based ventilation decisions
+  * Active climate control (beyond fans)
-  * Add active climate devices (beyond fans)
+    * Heater and/or active cooling
-    * Heater/ Active cooling
+    * Humidifier / dehumidifier system
-    * Humidifier / dehumidifier system
+  * Off-grid power system
-  * Off-grid power design outlook
+    * Battery/solar-based power supply for independent operation
-    * Plan and implement a battery/solar-based power supply so the system can operate independently of mains power.
+    * Power budgeting, electrical safety (e.g., fusing), and autonomy targets (e.g., nights or multiple cloudy days)
-    * Include power budgeting, safety (fusing), and autonomy targets (e.g., nights / multiple cloudy days).
 ===== 8. References / Sources =====