The objective of this study is to assess the efficacy of an Artificial Pancreas system which includes a Control-To-Range algorithm, versus standard Insulin Pump Therapy with Threshold Low glucose Suspend feature. 24 Children aged from 7 to 12 will be recruted in 4 clinical centres. After a 2 week run-in period, they will attend two investigational sessions of 65 hours in home-like settings. During each session, the CTR algorithm or the TLGS feature will be activated, depending on the randomization. The main endpoint is the time spent below 70mg/dL between 22:00 and 07:00.

The study is an open randomized control trial preceded by a run-in period and follows a cross-over design.After inclusion, the therapy will be moved to sensor augmented pump therapy with a dedicated training to continuous glucose monitoring using study system (study CGM). During 2 to 3 weeks, the patient will use his/her usual insulin pump and the study CGM (run-in period). At the end of this period, the patient and the relative will be admitted in a parents'house or hotel/resort nearby the hospital for 65 hours to undergo the first interventional session. Admission will be at 17:00 Day 1. The investigator will connect wirelessly the study pump and CGM to the study platform DiAs and remote monitoring will be activated. Until 08:00 Day 2, insulin delivery will be managed in open-loop mode, i.e. insulin doses will be decided according to capillary glucose measurements. From 08:00 Day 2, one algorithm will be activated according to the randomization order, i.e. either the threshold low glucose suspend (TLGS) algorithm (threshold will be set at 70 mg/dl) or the closed-loop algorithm. Insulin delivery will then be managed in open-loop mode with activated TLGS or by the closed-loop algorithm until 08:00 Day 4. In both cases, meals will be of free content and taken around 09:00, 13:00, 17:00 and 20:00. Meal insulin doses will be managed according to pre-meal glucose level and CHO meal content in open-loop mode with TLGS, or according to DiAs bolus calculator suggestion in closed-loop mode. Patient and relative will be encouraged to move in parents'house or hotel/resort and surroundings between meals. The algorithm will be inactivated at 08:00 Day 4. The study pump will be removed and replaced by patient's usual pump. The patient and the relative will be released at 10:00 after breakfast. Study CGM will still be used at home for 2 to 4 weeks until the second interventional session. This session will be similar to the first one but the alternative algorithm will be activated. Patient will come back to pre-study insulin therapy from 08:00 Day 4 and be released at 10:00 when second session is completed.

Inclusion Criteria:

1. Age from 7 to 12
2. Puberty status at Prader 1
3. Diabetes diagnosed according to WHO criteria since more than 1 year
4. Fasting plasma C-peptide level <0.2 ng/ml for a fasting blood glucose level < 180 mg/dl
5. At least one positiveness for plasma antibodies against pancreatic islets: IAA, IA2, GAD, ICA or ZnT8, since diagnosis of diabetes
6. Treatment of diabetes by insulin pump since more than 6 months
7. HbA1c level below 8.5%
8. Trained in carbohydrate counting
9. Lack of any associated disease or therapy (except insulin) affecting glucose metabolism
10. Willingness to follow all study procedures
11. Informed consent signed by patient and parent or legally responsible party
12. Patient must be affiliated or beneficiary of a social medical insurance

Exclusion Criteria:

1. Unwillingness of one parent or the legally responsible party to be present during all study procedures
2. Expected use of acetaminophen-containing drugs
3. Any disease or therapy (except insulin) affecting glucose metabolism during previous month
4. Impaired cognitive or psychological abilities which may result in defective adherence to study procedures
5. Active enrollment in another clinical trial or participation in a study within 30 days or participation in previous studies resulting in a cumulated annual income which would exceed 4500 €